Search results for: round-off error

Authors: B. R. Campomanes-Alvarez, P. Quiros, B. Fernandez

Abstract:

Automatic Speech Recognition (ASR) is a machine-based process of decoding and transcribing oral speech. A typical ASR system receives acoustic input from a speaker or an audio file, analyzes it using algorithms, and produces an output in the form of a text. Some speech recognition systems use Hidden Markov Models (HMMs) to deal with the temporal variability of speech and Gaussian Mixture Models (GMMs) to determine how well each state of each HMM fits a short window of frames of coefficients that represents the acoustic input. Another way to evaluate the fit is to use a feed-forward neural network that takes several frames of coefficients as input and produces posterior probabilities over HMM states as output. Deep neural networks (DNNs) that have many hidden layers and are trained using new methods have been shown to outperform GMMs on a variety of speech recognition systems. Acoustic models for state-of-the-art ASR systems are usually training on massive amounts of data. However, audio files with their corresponding transcriptions can be difficult to obtain, especially in the Spanish language. Hence, in the case of these low-resource scenarios, building an ASR model is considered as a complex task due to the lack of labeled data, resulting in an under-trained system. Semi-supervised learning approaches arise as necessary tasks given the high cost of transcribing audio data. The main goal of this proposal is to develop a procedure based on acoustic semi-supervised learning for Spanish ASR systems by using DNNs. This semi-supervised learning approach consists of: (a) Training a seed ASR model with a DNN using a set of audios and their respective transcriptions. A DNN with a one-hidden-layer network was initialized; increasing the number of hidden layers in training, to a five. A refinement, which consisted of the weight matrix plus bias term and a Stochastic Gradient Descent (SGD) training were also performed. The objective function was the cross-entropy criterion. (b) Decoding/testing a set of unlabeled data with the obtained seed model. (c) Selecting a suitable subset of the validated data to retrain the seed model, thereby improving its performance on the target test set. To choose the most precise transcriptions, three confidence scores or metrics, regarding the lattice concept (based on the graph cost, the acoustic cost and a combination of both), was performed as selection technique. The performance of the ASR system will be calculated by means of the Word Error Rate (WER). The test dataset was renewed in order to extract the new transcriptions added to the training dataset. Some experiments were carried out in order to select the best ASR results. A comparison between a GMM-based model without retraining and the DNN proposed system was also made under the same conditions. Results showed that the semi-supervised ASR-model based on DNNs outperformed the GMM-model, in terms of WER, in all tested cases. The best result obtained an improvement of 6% relative WER. Hence, these promising results suggest that the proposed technique could be suitable for building ASR models in low-resource environments.

Keywords: automatic speech recognition, deep neural networks, machine learning, semi-supervised learning

Procedia PDF Downloads 339

Authors: Lynn Sheridan, Dennis Alonzo, Hoa Nguyen, Andy Gao, Tracy Durksen

Abstract:

Disruptions to teaching (e.g., COVID-related) have increased work demands for teachers. There is an opportunity for research to explore evidence-informed steps to support teachers. Collective evidence informs data on teachers’ personal attributes (e.g., self-efficacy beliefs) in the workplace are seen to promote success in teaching and support teacher engagement. Teacher engagement plays a role in students’ learning and teachers’ effectiveness. Engaged teachers are better at overcoming work-related stress, burnout and are more likely to take on active roles. Teachers’ commitment is influenced by a host of personal (e.g., teacher well-being) and environmental factors (e.g., job stresses). The job demands-resources model provided a conceptual basis for examining how teachers’ well-being, and is influenced by job demands and job resources. Job demands potentially evoke strain and exceed the employee’s capability to adapt. Job resources entail what the job offers to individual teachers (e.g., organisational support), helping to reduce job demands. The application of the job demands-resources model involves gathering an evidence-base of and connection to personal attributes (job resources). The study explored the association between constructs (resilience, adaptability, commitment, self/collective efficacy) and a teacher’s engagement with the job. The paper sought to elaborate on the model and determine the associations between key constructs of well-being (resilience, adaptability), commitment, and motivation (self and collective-efficacy beliefs) to teachers’ engagement in teaching. Data collection involved online a multi-dimensional instrument using validated items distributed from 2020-2022. The instrument was designed to identify construct relationships. The participant number was 170. Data Analysis: The reliability coefficients, means, standard deviations, skewness, and kurtosis statistics for the six variables were completed. All scales have good reliability coefficients (.72-.96). A confirmatory factor analysis (CFA) and structural equation model (SEM) were performed to provide measurement support and to obtain latent correlations among factors. The final analysis was performed using structural equation modelling. Several fit indices were used to evaluate the model fit, including chi-square statistics and root mean square error of approximation. The CFA and SEM analysis was performed. The correlations of constructs indicated positive correlations exist, with the highest found between teacher engagement and resilience (r=.80) and the lowest between teacher adaptability and collective teacher efficacy (r=.22). Given the associations; we proceeded with CFA. The CFA yielded adequate fit: CFA fit: X (270, 1019) = 1836.79, p < .001, RMSEA = .04, and CFI = .94, TLI = .93 and SRMR = .04. All values were within the threshold values, indicating a good model fit. Results indicate that increasing teacher self-efficacy beliefs will increase a teacher’s level of engagement; that teacher ‘adaptability and resilience are positively associated with self-efficacy beliefs, as are collective teacher efficacy beliefs. Implications for school leaders and school systems: 1. investing in increasing teachers’ sense of efficacy beliefs to manage work demands; 2. leadership approaches can enhance teachers' adaptability and resilience; and 3. a culture of collective efficacy support. Preparing teachers for now and in the future offers an important reminder to policymakers and school leaders on the importance of supporting teachers’ personal attributes when faced with the challenging demands of the job.

Keywords: collective teacher efficacy, teacher self-efficacy, job demands, teacher engagement

Procedia PDF Downloads 124

Authors: Lucas L. L. Fortes, Sandro T. M. Gonçalves

Abstract:

In this work, it is analyzed the wideband performance with the mesh discretization impoverishment of the Conformal Finite Difference Time-Domain (C-FDTD) approaches developed by Raj Mittra, Supriyo Dey and Wenhua Yu for the Finite Difference Time-Domain (FDTD) method. These approaches are a simple and efficient way to optimize the scattering simulation of curved surfaces for Dielectric and Perfect Electric Conducting (PEC) structures in the FDTD method, since curved surfaces require dense meshes to reduce the error introduced due to the surface staircasing. Defined, on this work, as D-FDTD-Diel and D-FDTD-PEC, these approaches are well-known in the literature, but the improvement upon their application is not quantified broadly regarding wide frequency bands and poorly discretized meshes. Both approaches bring improvement of the accuracy of the simulation without requiring dense meshes, also making it possible to explore poorly discretized meshes which bring a reduction in simulation time and the computational expense while retaining a desired accuracy. However, their applications present limitations regarding the mesh impoverishment and the frequency range desired. Therefore, the goal of this work is to explore the approaches regarding both the wideband and mesh impoverishment performance to bring a wider insight over these aspects in FDTD applications. The D-FDTD-Diel approach consists in modifying the electric field update in the cells intersected by the dielectric surface, taking into account the amount of dielectric material within the mesh cells edges. By taking into account the intersections, the D-FDTD-Diel provides accuracy improvement at the cost of computational preprocessing, which is a fair trade-off, since the update modification is quite simple. Likewise, the D-FDTD-PEC approach consists in modifying the magnetic field update, taking into account the PEC curved surface intersections within the mesh cells and, considering a PEC structure in vacuum, the air portion that fills the intersected cells when updating the magnetic fields values. Also likewise to D-FDTD-Diel, the D-FDTD-PEC provides a better accuracy at the cost of computational preprocessing, although with a drawback of having to meet stability criterion requirements. The algorithms are formulated and applied to a PEC and a dielectric spherical scattering surface with meshes presenting different levels of discretization, with Polytetrafluoroethylene (PTFE) as the dielectric, being a very common material in coaxial cables and connectors for radiofrequency (RF) and wideband application. The accuracy of the algorithms is quantified, showing the approaches wideband performance drop along with the mesh impoverishment. The benefits in computational efficiency, simulation time and accuracy are also shown and discussed, according to the frequency range desired, showing that poorly discretized mesh FDTD simulations can be exploited more efficiently, retaining the desired accuracy. The results obtained provided a broader insight over the limitations in the application of the C-FDTD approaches in poorly discretized and wide frequency band simulations for Dielectric and PEC curved surfaces, which are not clearly defined or detailed in the literature and are, therefore, a novelty. These approaches are also expected to be applied in the modeling of curved RF components for wideband and high-speed communication devices in future works.

Keywords: accuracy, computational efficiency, finite difference time-domain, mesh impoverishment

Procedia PDF Downloads 134

Authors: Hiroshi Kinoshita, Yasuto Nakanishi, Ryuhei Okuno, Toshio Higashi

Abstract:

Various kinds of pegboards have been developed and used widely in research and clinics of rehabilitation for evaluation and training of patient’s hand function. A common measure in these peg boards is a total time of performance execution assessed by a tester’s stopwatch. Introduction of electrical and automatic measurement technology to the apparatus, on the other hand, has been delayed. The present work introduces the development of a pegboard with an electric sensor to detect moments of individual peg’s insertion and removal. The work also gives fundamental data obtained from a group of healthy young individuals who performed peg transfer tasks using the pegboard developed. Through trails and errors in pilot tests, two 10-hole peg-board boxes installed with a small photo-reflector and a DC amplifier at the bottom of each hole were designed and built by the present authors. The amplified electric analogue signals from the 20 reflectors were automatically digitized at 500 Hz per channel, and stored in a PC. The boxes were set on a test table at different distances (25, 50, 75, and 125 mm) in parallel to examine the effect of hole-to-hole distance. Fifty healthy young volunteers (25 in each gender) as subjects of the study performed successive fast 80 time peg transfers at each distance using their dominant and non-dominant hands. The data gathered showed a clear-cut light interruption/continuation moment by the pegs, allowing accurately (no tester’s error involved) and precisely (an order of milliseconds) to determine the pull out and insertion times of each peg. This further permitted computation of individual peg movement duration (PMD: from peg-lift-off to insertion) apart from hand reaching duration (HRD: from peg insertion to lift-off). An accidental drop of a peg led to an exceptionally long ( < mean + 3 SD) PMD, which was readily detected from an examination of data distribution. The PMD data were commonly right-skewed, suggesting that the median can be a better estimate of individual PMD than the mean. Repeated measures ANOVA using the median values revealed significant hole-to-hole distance, and hand dominance effects, suggesting that these need to be fixed in the accurate evaluation of PMD. The gender effect was non-significant. Performance consistency was also evaluated by the use of quartile variation coefficient values, which revealed no gender, hole-to-hole, and hand dominance effects. The measurement reliability was further examined using interclass correlation obtained from 14 subjects who performed the 25 and 125 mm hole distance tasks at two 7-10 days separate test sessions. Inter-class correlation values between the two tests showed fair reliability for PMD (0.65-0.75), and for HRD (0.77-0.94). We concluded that a sensor peg board developed in the present study could provide accurate (excluding tester’s errors), and precise (at a millisecond rate) time information of peg movement separated from that used for hand movement. It could also easily detect and automatically exclude erroneous execution data from his/her standard data. These would lead to a better evaluation of hand dexterity function compared to the widely used conventional used peg boards.

Keywords: hand, dexterity test, peg movement time, performance consistency

Procedia PDF Downloads 133

Authors: Rifilwe Victor Modiba, Stefan Hendric Foord

Abstract:

Alien invasive plants (IAP’s) have considerable negative impacts on freshwater habitats and South Africa has implemented an innovative Work for Water (WfW) programme for the systematic removal of these plants aimed at, amongst other objectives, restoring biodiversity and ecosystem services in these threatened habitats. These restoration processes are expensive and have to be evidence-based. In this study in-stream macroinvertebrate and adult Odonata assemblages were used as indicators of restoration success by quantifying the response of biodiversity metrics for these two groups to the removal of IAP’s in a strategic water resource of South Africa that is extensively invaded by invasive alien plants (IAP’s). The study consisted of a replicated design that included 45 sampling units, viz. 15 invaded, 15 uninvaded and 15 cleared sites stratified across the upper reaches of six sub-catchments of the Luvuvhu river catchment, Limpopo Province. Cleared sites were only considered if they received at least two WfW treatments in the last 3 years. The Benthic macroinvertebrate and adult Odonate assemblages in each of these sampling were surveyed from between November and March, 2013/2014 and 2014/2015 respectively. Generalized Linear Models (GLM) with a log link function and Poisson error distribution were done for metrics (invaded, cleared, and uninvaded) whose residuals were not normally distributed or had unequal variance and for abundance. RDA was done for EPTO genera (Ephemeroptera, Plecoptera, Trichoptera and Odonata) and adult Odonata species abundance. GLM was done to for the abundance of Genera and Odonates that had the association with the RDA environmental factors. Sixty four benthic macroinvertebrate families, 57 EPTO genera, and 45 adult Odonata species were recorded across all 45 sampling units. There was no significant difference between the SASS5 total score, ASPT, and family richness of the three invasion classes. Although clearing only had a weak positive effect on the adult Odonate species richness it had a positive impact on DBI scores. These differences were mainly the result of significantly larger DBI scores in the cleared sites as compared to the invaded sites. Results suggest that water quality is positively impacted by repeated clearing pointing to the importance of follow up procedures after initial clearing. Adult Odonate diversity as measured by richness, endemicity, threat and distribution respond positively to all forms of the clearing. The clearing had a significant impact on Odonate assemblage structure but did not affect EPTO structure. Variation partitioning showed that 21.8% of the variation in EPTO assemblage can be explained by spatial and environmental variables, 16% of the variation in Odonate structure was explained by spatial and environmental variables. The response of the diversity metrics to clearing increased in significance at finer taxonomic resolutions, particularly of adult Odonates whose metrics significantly improved with clearing and whose structure responded to both invasion and clearing. The study recommends the use of DBI for surveying river health when hydraulic biotopes are poor.

Keywords: DBI, evidence-based conservation, EPTO, macroinvetebrates

Procedia PDF Downloads 186

Authors: Jungyeol Hong, Dongjoo Park

Abstract:

The primary purpose of this study is to develop a methodological framework to predict daily subway ridership at a station-level and to examine the association between subway ridership and bus demand incorporating commercial business facility in the vicinity of each subway station. The socio-economic characteristics, land-use, and built environment as factors may have an impact on subway ridership. However, it should be considered not only the endogenous relationship between bus and subway demand but also the characteristics of commercial business within a subway station’s sphere of influence, and integrated transit network topology. Regarding a statistical approach to estimate subway ridership at a station level, therefore it should be considered endogeneity and heteroscedastic issues which might have in the subway ridership prediction model. This study focused on both discovering the impacts of bus demand, commercial business characteristics, and network topology on subway ridership and developing more precise subway ridership estimation accounting for its statistical bias. The spatial scope of the study covers entire Seoul city in South Korea and includes 243 stations with the temporal scope set at twenty-four hours with one-hour interval time panels each. The data for subway and bus ridership was collected Seoul Smart Card data from 2015 and 2016. Three-Stage Least Square(3SLS) approach was applied to develop daily subway ridership model as capturing the endogeneity and heteroscedasticity between bus and subway demand. Independent variables incorporating in the modeling process were commercial business characteristics, social-economic characteristics, safety index, transit facility attributes, and dummies for seasons and time zone. As a result, it was found that bus ridership and subway ridership were endogenous each other and they had a significantly positive sign of coefficients which means one transit mode could increase another transportation mode’s ridership. In other words, two transit modes of subway and bus have a mutual relationship instead of the competitive relationship. The commercial business characteristics are the most critical dimension among the independent variables. The variables of commercial business facility rate in the paper containing six types; medical, educational, recreational, financial, food service, and shopping. From the model result, a higher rate in medical, financial buildings, shopping, and food service facility lead to increment of subway ridership at a station, while recreational and educational facility shows lower subway ridership. The complex network theory was applied for estimating integrated network topology measures that cover the entire Seoul transit network system, and a framework for seeking an impact on subway ridership. The centrality measures were found to be significant and showed a positive sign indicating higher centrality led to more subway ridership at a station level. The results of model accuracy tests by out of samples provided that 3SLS model has less mean square error rather than OLS and showed the methodological approach for the 3SLS model was plausible to estimate more accurate subway ridership. Acknowledgement: This research was supported by Basic Science Research Program through the National Research Foundation of Korea(NRF) funded by the Ministry of Science and ICT (2017R1C1B2010175).

Keywords: subway ridership, bus ridership, commercial business characteristic, endogeneity, network topology

Procedia PDF Downloads 144

Authors: Mingxin Li, Liya Ni

Abstract:

Autonomous parking is a valuable feature applicable to many robotics applications such as tour guide robots, UV sanitizing robots, food delivery robots, and warehouse robots. With auto-parking, the robot will be able to park at the charging zone and charge itself without human intervention. As compared to self-driving vehicles, auto-parking is more challenging for a small-scale mobile robot only equipped with a front camera due to the camera view limited by the robot’s height and the narrow Field of View (FOV) of the inexpensive camera. In this research, auto-parking of a small-scale mobile robot with a front camera only was achieved in a four-step process: Firstly, transfer learning was performed on the AlexNet, a popular pre-trained convolutional neural network (CNN). It was trained with 150 pictures of empty parking slots and 150 pictures of occupied parking slots from the view angle of a small-scale robot. The dataset of images was divided into a group of 70% images for training and the remaining 30% images for validation. An average success rate of 95% was achieved. Secondly, the image of detected empty parking space was processed with edge detection followed by the computation of parametric representations of the boundary lines using the Hough Transform algorithm. Thirdly, the positions of the entrance point and center of available parking space were predicted based on the robot kinematic model as the robot was driving closer to the parking space because the boundary lines disappeared partially or completely from its camera view due to the height and FOV limitations. The robot used its wheel speeds to compute the positions of the parking space with respect to its changing local frame as it moved along, based on its kinematic model. Lastly, the predicted entrance point of the parking space was used as the reference for the motion control of the robot until it was replaced by the actual center when it became visible again by the robot. The linear and angular velocities of the robot chassis center were computed based on the error between the current chassis center and the reference point. Then the left and right wheel speeds were obtained using inverse kinematics and sent to the motor driver. The above-mentioned four subtasks were all successfully accomplished, with the transformed learning, image processing, and target prediction performed in MATLAB, while the motion control and image capture conducted on a self-built small scale differential drive mobile robot. The small-scale robot employs a Raspberry Pi board, a Pi camera, an L298N dual H-bridge motor driver, a USB power module, a power bank, four wheels, and a chassis. Future research includes three areas: the integration of all four subsystems into one hardware/software platform with the upgrade to an Nvidia Jetson Nano board that provides superior performance for deep learning and image processing; more testing and validation on the identification of available parking space and its boundary lines; improvement of performance after the hardware/software integration is completed.

Keywords: autonomous parking, convolutional neural network, image processing, kinematics-based prediction, transfer learning

Procedia PDF Downloads 132

Authors: D. Madhushanka, Y. Liu, H. C. Fernando

Abstract:

Fires are one of the foremost factors of land surface disturbance in diverse ecosystems, causing soil erosion and land-cover changes and atmospheric effects affecting people's lives and properties. Generally, the severity of the fire is calculated as the Normalized Burn Ratio (NBR) index. This is performed manually by comparing two images obtained afterward. Then by using the bitemporal difference of the preprocessed satellite images, the dNBR is calculated. The burnt area is then classified as either unburnt (dNBR<0.1) or burnt (dNBR>= 0.1). Furthermore, Wildfire Severity Assessment (WSA) classifies burnt areas and unburnt areas using classification levels proposed by USGS and comprises seven classes. This procedure generates a burn severity report for the area chosen by the user manually. This study is carried out with the objective of producing an automated tool for the above-mentioned process, namely the World Wildfire Severity Assessment Tool (WWSAT). It is implemented in Google Earth Engine (GEE), which is a free cloud-computing platform for satellite data processing, with several data catalogs at different resolutions (notably Landsat, Sentinel-2, and MODIS) and planetary-scale analysis capabilities. Sentinel-2 MSI is chosen to obtain regular processes related to burnt area severity mapping using a medium spatial resolution sensor (15m). This tool uses machine learning classification techniques to identify burnt areas using NBR and to classify their severity over the user-selected extent and period automatically. Cloud coverage is one of the biggest concerns when fire severity mapping is performed. In WWSAT based on GEE, we present a fully automatic workflow to aggregate cloud-free Sentinel-2 images for both pre-fire and post-fire image compositing. The parallel processing capabilities and preloaded geospatial datasets of GEE facilitated the production of this tool. This tool consists of a Graphical User Interface (GUI) to make it user-friendly. The advantage of this tool is the ability to obtain burn area severity over a large extent and more extended temporal periods. Two case studies were carried out to demonstrate the performance of this tool. The Blue Mountain national park forest affected by the Australian fire season between 2019 and 2020 is used to describe the workflow of the WWSAT. This site detected more than 7809 km2, using Sentinel-2 data, giving an error below 6.5% when compared with the area detected on the field. Furthermore, 86.77% of the detected area was recognized as fully burnt out, of which high severity (17.29%), moderate-high severity (19.63%), moderate-low severity (22.35%), and low severity (27.51%). The Arapaho and Roosevelt National Forest Park, California, the USA, which is affected by the Cameron peak fire in 2020, is chosen for the second case study. It was found that around 983 km2 had burned out, of which high severity (2.73%), moderate-high severity (1.57%), moderate-low severity (1.18%), and low severity (5.45%). These spots also can be detected through the visual inspection made possible by cloud-free images generated by WWSAT. This tool is cost-effective in calculating the burnt area since satellite images are free and the cost of field surveys is avoided.

Keywords: burnt area, burnt severity, fires, google earth engine (GEE), sentinel-2

Procedia PDF Downloads 235

Authors: Ariunaa.S., Javzandulam E., Chimegsaikhan S., Altantsetseg B., Oyungerel S., Bat-Erdene T., Naranbaatar S., Otgonbayar B., Suvdaa N., Tumenbayar B.

Abstract:

Introduction: Prenatal stress has been linked to heightened allergy sensitivity in offspring. However, there is a notable absence of research on the mesovarium structure of offspring born from mothers subjected to cold stress during pregnancy. Understanding the impact of maternal cold stress on the mesovarium structure could provide valuable insights into reproductive health outcomes in offspring. Objective: This study aims to investigate structural changes in the mesovarium of offspring born from cold-stress affected rats. Material and Methods: 20 female Westar rats weighing around 200g were chosen and evenly divided into four containers; then, 2-3 male rats were introduced to each container. The Papanicolaou method was used to estimate the spermatozoa and estrus period from vaginal swabs taken from female rats at 8:00 a.m. Female rats examined with the presence of spermatozoa during the estrous phase of the estrous cycle are defined as pregnant. Pregnant rats are divided into experimental and control groups. The experimental group was stressed using the model of severe and chronic cold stress for 30 days. They were exposed to cold stress for 3 hours each morning between 8:00 and 11:00 o’clock at a temperature of minus 15 degrees Celsius. The control group was kept under normal laboratory conditions. Newborn female rats from both experimental and control groups were selected. At 2 months of age, rats were euthanized by decapitation, and their mesovaria were collected. Tissues were fixed in 4% formalin, embedded in paraffin, and sectioned into 5μm thick slices. The sections were stained with H&E and digitized by digital microscope. The area of brown fat and inflammatory infiltrations were quantified using Image J software. The blood cortisol levels were measured using ELISA. Data are expressed as the mean ± standard error of the mean (SEM). The Mann-Whitney test was used to compare the two groups. All analyses were performed using Prism (GraphPad Software). A p-value of < 0.05 was considered statistically significant. Result: Offspring born from stressed mothers exhibited significant physiological differences compared to the control group. Specifically, the body weight of offspring from stressed mothers was significantly lower than the control group (p=0.0002). Conversely, the cortisol level in offspring from stressed mothers was significantly higher (p=0.0446). Offspring born from stressed mothers showed a statistically significant increase in brown fat area compared to the control group (p=0.01). Additionally, offspring from stressed mothers had a significantly higher number of inflammatory infiltrates in their mesovarium compared to the control group (p<0.047). These results indicate the profound impact of maternal stress on offspring physiology, affecting body weight, stress hormone levels, metabolic characteristics, and inflammatory responses. Conclusion: Exposure to cold stress during pregnancy has significant repercussions on offspring physiology. Our findings demonstrate that cold stress exposure leads to increased blood cortisol levels, brown fat accumulation, and inflammatory cell infiltration in offspring. These results underscore the profound impact of maternal stress on offspring health and highlight the importance of mitigating environmental stressors during pregnancy to promote optimal offspring outcomes.

Keywords: brown fat, cold stress during pregnancy, inflammation, mesovarium

Procedia PDF Downloads 45

Authors: Swati Bhomia, C. M. Kishtawal, Neeru Jaiswal

Abstract:

Traditionally, monsoon forecasts have encountered many difficulties that stem from numerous issues such as lack of adequate upper air observations, mesoscale nature of convection, proper resolution, radiative interactions, planetary boundary layer physics, mesoscale air-sea fluxes, representation of orography, etc. Uncertainties in any of these areas lead to large systematic errors. Global circulation models (GCMs), which are developed independently at different institutes, each of which carries somewhat different representation of the above processes, can be combined to reduce the collective local biases in space, time, and for different variables from different models. This is the basic concept behind the multi-model superensemble and comprises of a training and a forecast phase. The training phase learns from the recent past performances of models and is used to determine statistical weights from a least square minimization via a simple multiple regression. These weights are then used in the forecast phase. The superensemble forecasts carry the highest skill compared to simple ensemble mean, bias corrected ensemble mean and the best model out of the participating member models. This approach is a powerful post-processing method for the estimation of weather forecast parameters reducing the direct model output errors. Although it can be applied successfully to the continuous parameters like temperature, humidity, wind speed, mean sea level pressure etc., in this paper, this approach is applied to rainfall, a parameter quite difficult to handle with standard post-processing methods, due to its high temporal and spatial variability. The present study aims at the development of advanced superensemble schemes comprising of 1-5 day daily precipitation forecasts from five state-of-the-art global circulation models (GCMs), i.e., European Centre for Medium Range Weather Forecasts (Europe), National Center for Environmental Prediction (USA), China Meteorological Administration (China), Canadian Meteorological Centre (Canada) and U.K. Meteorological Office (U.K.) obtained from THORPEX Interactive Grand Global Ensemble (TIGGE), which is one of the most complete data set available. The novel approaches include the dynamical model selection approach in which the selection of the superior models from the participating member models at each grid and for each forecast step in the training period is carried out. Multi-model superensemble based on the training using similar conditions is also discussed in the present study, which is based on the assumption that training with the similar type of conditions may provide the better forecasts in spite of the sequential training which is being used in the conventional multi-model ensemble (MME) approaches. Further, a variety of methods that incorporate a 'neighborhood' around each grid point which is available in literature to allow for spatial error or uncertainty, have also been experimented with the above mentioned approaches. The comparison of these schemes with respect to the observations verifies that the newly developed approaches provide more unified and skillful prediction of the summer monsoon (viz. June to September) rainfall compared to the conventional multi-model approach and the member models.

Keywords: multi-model superensemble, dynamical model selection, similarity criteria, neighborhood technique, rainfall prediction

Procedia PDF Downloads 139

Authors: Avdhesh K. Sharma

Abstract:

Key, in the development of energy-efficient micro-scale heat exchanger devices, is to select large heat transfer surface to volume ratio without much expanse on re-circulated pumps. The increased interest in short heat exchanger (SHE) is due to accessibility of advanced technologies for manufacturing of micro-tubes in range of 1 micron m - 1 mm. Such SHE using micro-tubes are highly effective for high flux heat transfer technologies. Nanofluids, are used to enhance the thermal conductivity of re-circulated coolant and thus enhances heat transfer rate further. Higher viscosity associated with nanofluid expands more pumping power. Thus, there is a trade-off between heat transfer rate and pressure drop with geometry of micro-tubes. Herein, a novel design of short counter flow heat exchanger (SCFHE) using non-circular micro-tubes flooded with CuO-water nanofluid is conceptualized by varying the ratio of surface area to cross-sectional area of micro-tubes. A framework for comparative analysis of SCFHE using micro-tubes non-circular shape flooded by CuO-water nanofluid is presented. In SCFHE concept, micro-tubes having various geometrical shapes (viz., triangular, rectangular and trapezoidal) has been arranged row-wise to facilitate two aspects: (1) allowing easy flow distribution for cold and hot stream, and (2) maximizing the thermal interactions with neighboring channels. Adequate distribution of rows for cold and hot flow streams enables above two aspects. For comparative analysis, a specific volume or cross-section area is assigned to each elemental cell (which includes flow area and area corresponds to half wall thickness). A specific volume or cross-section area is assumed to be constant for each elemental cell (which includes flow area and half wall thickness area) and variation in surface area is allowed by selecting different geometry of micro-tubes in SCFHE. Effective thermal conductivity model for CuO-water nanofluid has been adopted, while the viscosity values for water based nanofluids are obtained empirically. Correlations for Nusselt number (Nu) and Poiseuille number (Po) for micro-tubes have been derived or adopted. Entrance effect is accounted for. Thermal and hydrodynamic performances of SCFHE are defined in terms of effectiveness and pressure drop or pumping power, respectively. For defining the overall performance index of SCFHE, two links are employed. First one relates heat transfer between the fluid streams q and pumping power PP as (=qj/PPj); while another link relates effectiveness eff and pressure drop dP as (=effj/dPj). For analysis, the inlet temperatures of hot and cold streams are varied in usual range of 20dC-65dC. Fully turbulent regime is seldom encountered in micro-tubes and transition of flow regime occurs much early (i.e., ~Re=1000). Thus, Re is fixed at 900, however, the uncertainty in Re due to addition of nanoparticles in base fluid is quantified by averaging of Re. Moreover, for minimizing error, volumetric concentration is limited to range 0% to ≤4% only. Such framework may be helpful in utilizing maximum peripheral surface area of SCFHE without any serious severity on pumping power and towards developing advanced short heat exchangers.

Keywords: CuO-water nanofluid, non-circular micro-tubes, performance index, short counter flow heat exchanger

Procedia PDF Downloads 213

Authors: Neil Singla, Derek Muse, Karen DiDonato, Pamela Palmer

Abstract:

Introduction: Pain is the most common reason people visit emergency rooms. Studies indicate however, that Emergency Department (ED) physicians often do not provide adequate analgesia to their patients as a result of gender and age bias, opiophobia and insufficient knowledge of and formal training in acute pain management. Novel classes of analgesics have recently been introduced, but many patients suffer from acute pain in settings where the availability of intravenous (IV) access may be limited, so there remains a clinical need for rapid-acting, potent analgesics that do not require an invasive route of delivery. A sublingual sufentanil tablet (SST), dispensed using a single-dose applicator, is in development for treatment of moderate-to-severe acute pain in a medically-supervised setting. Objective: The primary objective of this study was to demonstrate the repeat-dose efficacy, safety and tolerability of sufentanil 20 mcg and 30 mcg sublingual tablets compared to placebo for the management of acute pain as determined by the time-weighted sum of pain intensity differences (SPID) to baseline over the 12-hour study period (SPID12). Key secondary efficacy variables included SPID over the first hour (SPID1), Total pain relief over the 12-hour study period (TOTPAR12), time to perceived pain relief (PR) and time to meaningful PR. Safety variables consisted of adverse events (AE), vital signs, oxygen saturation and early termination. Methods: In this Phase 2, double-blind, dose-finding study, an equal number of male and female patients were randomly assigned in a 2:2:1 ratio to SST 20 mcg, SS 30 mcg or placebo, respectively, following bunionectomy. Study drug was dosed as needed, but not more frequently than hourly. Rescue medication was available as needed. The primary endpoint was the Summed Pain Intensity Difference to baseline over 12h (SPIDI2). Safety was assessed by continuous oxygen saturation monitoring and adverse event reporting. Results: 101 patients (51 Male/50 Female) were randomized, 100 received study treatment (intent-to-treat [ITT] population), and 91 completed the study. Reasons for early discontinuation were lack of efficacy (6), adverse events (2) and drug-dosing error (1). Mean age was 42.5 years. For the ITT population, SST 30 mcg was superior to placebo (p=0.003) for the SPID12. SPID12 scores in the active groups were superior for both male (ANOVA overall p-value =0.038) and female (ANOVA overall p-value=0.005) patients. Statistically significant differences in favour of sublingual sufentanil were also observed between the SST 30mcg and placebo group for SPID1(p<0.001), TOTPAR12(p=0.002), time to perceived PR (p=0.023) and time to meaningful PR (p=0.010). Nausea, vomiting and somnolence were more frequent in the sufentanil groups but there were no significant differences between treatment arms for the proportion of patients who prematurely terminated due to AE or inadequate analgesia. Conclusions: Sufentanil tablets dispensed sublingually using a single-dose applicator is in development for treatment of patients with moderate-to-severe acute pain in a medically-supervised setting where immediate IV access is limited. When administered sublingually, sufentanil’s pharmacokinetic profile and non-invasive delivery makes it a useful alternative to IM or IV dosing.

Keywords: acute pain, pain management, sublingual, sufentanil

Procedia PDF Downloads 356

Authors: Parth Gupta, Ujjawal Singh, Shashank Kumar, Mansi Chandra, Arnab Sarkar

Abstract:

Microfluidic devices have emerged as indispensable tools across various scientific disciplines, offering precise control and manipulation of fluids at the microscale. Their efficacy in flow-based research, spanning engineering, chemistry, and biology, relies heavily on the geometric design of microfluidic channels. This work introduces a novel approach to optimise these channels through Response Surface Methodology (RSM), departing from the conventional practice of addressing one parameter at a time. Traditionally, optimising microfluidic channels involved isolated adjustments to individual parameters, limiting the comprehensive understanding of their combined effects. In contrast, our approach considers the simultaneous impact of multiple parameters, employing RSM to efficiently explore the complex design space. The outcome is an innovative microfluidic channel that consumes an optimal sample volume and minimises flow time, enhancing overall efficiency. The relevance of geometric parameter optimization in microfluidic channels extends significantly in biomedical engineering. The flow characteristics of porous materials within these channels depend on many factors, including fluid viscosity, environmental conditions (such as temperature and humidity), and specific design parameters like sample volume, channel width, channel length, and substrate porosity. This intricate interplay directly influences the performance and efficacy of microfluidic devices, which, if not optimized, can lead to increased costs and errors in disease testing and analysis. In the context of biomedical applications, the proposed approach addresses the critical need for precision in fluid flow. it mitigate manufacturing costs associated with trial-and-error methodologies by optimising multiple geometric parameters concurrently. The resulting microfluidic channels offer enhanced performance and contribute to a streamlined, cost-effective process for testing and analyzing diseases. A key highlight of our methodology is its consideration of the interconnected nature of geometric parameters. For instance, the volume of the sample, when optimized alongside channel width, length, and substrate porosity, creates a synergistic effect that minimizes errors and maximizes efficiency. This holistic optimization approach ensures that microfluidic devices operate at their peak performance, delivering reliable results in disease testing. A key highlight of our methodology is its consideration of the interconnected nature of geometric parameters. For instance, the volume of the sample, when optimized alongside channel width, length, and substrate porosity, creates a synergistic effect that minimizes errors and maximizes efficiency. This holistic optimization approach ensures that microfluidic devices operate at their peak performance, delivering reliable results in disease testing. A key highlight of our methodology is its consideration of the interconnected nature of geometric parameters. For instance, the volume of the sample, when optimized alongside channel width, length, and substrate porosity, creates a synergistic effect that minimizes errors and maximizes efficiency. This holistic optimization approach ensures that microfluidic devices operate at their peak performance, delivering reliable results in disease testing.

Keywords: microfluidic device, minitab, statistical optimization, response surface methodology

Procedia PDF Downloads 68

Authors: Pablo Chico De Guzman, Cesar Sanchez

Abstract:

Microservice architectures are increasingly in distributed cloud applications due to the advantages on the software composition, development speed, release cycle frequency and the business logic time to market. On the other hand, these architectures also introduce some challenges on the testing and release phases of applications. Container technology solves some of these issues by providing reproducible environments, easy of software distribution and isolation of processes. However, there are other issues that remain unsolved in current container technology when dealing with multiple machines, such as networking for multi-host communication, service discovery, load balancing or data persistency (even though some of these challenges are already solved by traditional cloud vendors in a very mature and widespread manner). Container cluster management tools, such as Kubernetes, Mesos or Docker Swarm, attempt to solve these problems by introducing a new control layer where the unit of deployment is the container (or the pod — a set of strongly related containers that must be deployed on the same machine). These tools are complex to configure and manage and they do not follow a pure immutable infrastructure approach since servers are reused between deployments. Indeed, these tools introduce dependencies at execution time for solving networking or service discovery problems. If an error on the control layer occurs, which would affect running applications, specific expertise is required to perform ad-hoc troubleshooting. As a consequence, it is not surprising that container cluster support is becoming a source of revenue for consulting services. This paper presents i2kit, a deployment tool based on the immutable infrastructure pattern, where the virtual machine is the unit of deployment. The input for i2kit is a declarative definition of a set of microservices, where each microservice is defined as a pod of containers. Microservices are built into machine images using linuxkit —- a tool for creating minimal linux distributions specialized in running containers. These machine images are then deployed to one or more virtual machines, which are exposed through a cloud vendor load balancer. Finally, the load balancer endpoint is set into other microservices using an environment variable, providing service discovery. The toolkit i2kit reuses the best ideas from container technology to solve problems like reproducible environments, process isolation, and software distribution, and at the same time relies on mature, proven cloud vendor technology for networking, load balancing and persistency. The result is a more robust system with no learning curve for troubleshooting running applications. We have implemented an open source prototype that transforms i2kit definitions into AWS cloud formation templates, where each microservice AMI (Amazon Machine Image) is created on the fly using linuxkit. Even though container cluster management tools have more flexibility for resource allocation optimization, we defend that adding a new control layer implies more important disadvantages. Resource allocation is greatly improved by using linuxkit, which introduces a very small footprint (around 35MB). Also, the system is more secure since linuxkit installs the minimum set of dependencies to run containers. The toolkit i2kit is currently under development at the IMDEA Software Institute.

Keywords: container, deployment, immutable infrastructure, microservice

Procedia PDF Downloads 179

Authors: M. Reza Hashemi, Chris Small, Scott Hayward

Abstract:

The Northeast Coast of the US faces damaging effects of coastal flooding and winds due to Atlantic tropical and extratropical storms each year. Historically, several large storm events have produced substantial levels of damage to the region; most notably of which were the Great Atlantic Hurricane of 1938, Hurricane Carol, Hurricane Bob, and recently Hurricane Sandy (2012). The objective of this study was to develop an integrated modeling system that could be used as a forecasting/hindcasting tool to evaluate and communicate the risk coastal communities face from these coastal storms. This modeling system utilizes the ADvanced CIRCulation (ADCIRC) model for storm surge predictions and the Simulating Waves Nearshore (SWAN) model for the wave environment. These models were coupled, passing information to each other and computing over the same unstructured domain, allowing for the most accurate representation of the physical storm processes. The coupled SWAN-ADCIRC model was validated and has been set up to perform real-time forecast simulations (as well as hindcast). Modeled storm parameters were then passed to a coastal risk assessment tool. This tool, which is generic and universally applicable, generates spatial structural damage estimate maps on an individual structure basis for an area of interest. The required inputs for the coastal risk model included a detailed information about the individual structures, inundation levels, and wave heights for the selected region. Additionally, calculation of wind damage to structures was incorporated. The integrated coastal risk assessment system was then tested and applied to Charlestown, a small vulnerable coastal town along the southern shore of Rhode Island. The modeling system was applied to Hurricane Sandy and a synthetic storm. In both storm cases, effect of natural dunes on coastal risk was investigated. The resulting damage maps for the area (Charlestown) clearly showed that the dune eroded scenarios affected more structures, and increased the estimated damage. The system was also tested in forecast mode for a large Nor’Easters: Stella (March 2017). The results showed a good performance of the coupled model in forecast mode when compared to observations. Finally, a nearshore model XBeach was then nested within this regional grid (ADCIRC-SWAN) to simulate nearshore sediment transport processes and coastal erosion. Hurricane Irene (2011) was used to validate XBeach, on the basis of a unique beach profile dataset at the region. XBeach showed a relatively good performance, being able to estimate eroded volumes along the beach transects with a mean error of 16%. The validated model was then used to analyze the effectiveness of several erosion mitigation methods that were recommended in a recent study of coastal erosion in New England: beach nourishment, coastal bank (engineered core), and submerged breakwater as well as artificial surfing reef. It was shown that beach nourishment and coastal banks perform better to mitigate shoreline retreat and coastal erosion.

Keywords: ADCIRC, coastal flooding, storm surge, coastal risk assessment, living shorelines

Procedia PDF Downloads 116

Authors: Lionel Trovalet, Lisa Liu, Dimitri Bigot, Nadia Hammami, Jean-Pierre Habas, Bruno Malet-Damour

Abstract:

The MCP-iBAT1 project is carried out to study the behavior of Phase Change Materials (PCM) integrated in building envelopes in a tropical environment. Through the phase transitions (melting and freezing) of the material, thermal energy can be absorbed or released. This process enables the regulation of indoor temperatures and the improvement of thermal comfort for the occupants. Most of the commercially available PCMs are more suitable to temperate climates than to tropical climates. The case of Reunion Island is noteworthy as there are multiple micro-climates. This leads to our key question: developing one or multiple bio-based PCMs that cover the thermal needs of the different locations of the island. The present paper focuses on the numerical approach to select the PCM properties relevant to tropical areas. Numerical simulations have been carried out with two softwares: EnergyPlusTM and Isolab. The latter has been developed in the laboratory, with the implicit Finite Difference Method, in order to evaluate different physical models. Both are Thermal Dynamic Simulation (TDS) softwares that predict the building’s thermal behavior with one-dimensional heat transfers. The parameters used in this study are the construction’s characteristics (dimensions and materials) and the environment’s description (meteorological data and building surroundings). The building is modeled in accordance with the experimental setup. It is divided into two rooms, cells A and B, with same dimensions. Cell A is the reference, while in cell B, a layer of commercial PCM (Thermo Confort of MCI Technologies) has been applied to the inner surface of the North wall. Sensors are installed in each room to retrieve temperatures, heat flows, and humidity rates. The collected data are used for the comparison with the numerical results. Our strategy is to implement two similar buildings at different altitudes (Saint-Pierre: 70m and Le Tampon: 520m) to measure different temperature ranges. Therefore, we are able to collect data for various seasons during a condensed time period. The following methodology is used to validate the numerical models: calibration of the thermal and PCM models in EnergyPlusTM and Isolab based on experimental measures, then numerical testing with a sensitivity analysis of the parameters to reach the targeted indoor temperatures. The calibration relies on the past ten months’ measures (from September 2020 to June 2021), with a focus on one-week study on November (beginning of summer) when the effect of PCM on inner surface temperatures is more visible. A first simulation with the PCM model of EnergyPlus gave results approaching the measurements with a mean error of 5%. The studied property in this paper is the melting temperature of the PCM. By determining the representative temperature of winter, summer and inter-seasons with past annual’s weather data, it is possible to build a numerical model of multi-layered PCM. Hence, the combined properties of the materials will provide an optimal scenario for the application on PCM in tropical areas. Future works will focus on the development of bio-based PCMs with the selected properties followed by experimental and numerical validation of the materials. 1Materiaux ´ a Changement de Phase, une innovation pour le B ` ati Tropical

Keywords: energyplus, multi-layer of PCM, phase changing materials, tropical area

Procedia PDF Downloads 95

Authors: Judy M. Obliosca, Olivia Vest, Sandra Poulos, Kelsi Smith, Tammy Ferguson, Abigail Powers Lott, Alicia K. Smith, Yang Xu, Christopher K. Tison

Abstract:

Post-Traumatic Stress Disorder (PTSD) is a mental health problem that people may develop after experiencing traumatic events such as combat, natural disasters, and major emotional challenges. Tragically, the number of military personnel with PTSD correlates directly with the number of veterans who attempt suicide, with the highest rate in the Army. Research has shown epigenetic risks in those who are prone to several psychiatric dysfunctions, particularly PTSD. Once initiated in response to trauma, epigenetic alterations in particular, the DNA methylation in the form of 5-methylcytosine (5mC) alters chromatin structure and represses gene expression. Current methods to detect DNA methylation, such as bisulfite-based genomic sequencing techniques, are laborious and have massive analysis workflow while still having high error rates. A faster and simpler detection method of high sensitivity and precision would be useful in a clinical setting to confirm potential PTSD etiologies, prevent other psychiatric disorders, and improve military health. A nano-enhanced Surface Plasmon Resonance imaging (SPRi)-based assay that simultaneously detects site-specific 5mC base (termed as PTSD base) in methylated genes related to PTSD is being developed. The arrays on a sensing chip were first constructed for parallel detection of PTSD bases using synthetic and genomic DNA (gDNA) samples. For the gDNA sample extracted from the whole blood of a PTSD patient, the sample was first digested using specific restriction enzymes, and fragments were denatured to obtain single-stranded methylated target genes (ssDNA). The resulting mixture of ssDNA was then injected into the assay platform, where targets were captured by specific DNA aptamer probes previously immobilized on the surface of a sensing chip. The PTSD bases in targets were detected by anti-5-methylcytosine antibody (anti-5mC), and the resulting signals were then enhanced by the universal nanoenhancer. Preliminary results showed successful detection of a PTSD base in a gDNA sample. Brighter spot images and higher delta values (control-subtracted reflectivity signal) relative to those of the control were observed. We also implemented the in-house surface activation system for detection and developed SPRi disposable chips. Multiplexed PTSD base detection of target methylated genes in blood DNA from PTSD patients of severity conditions (asymptomatic and severe) was conducted. This diagnostic capability being developed is a platform technology, and upon successful implementation for PTSD, it could be reconfigured for the study of a wide variety of neurological disorders such as traumatic brain injury, Alzheimer’s disease, schizophrenia, and Huntington's disease and can be extended to the analyses of other sample matrices such as urine and saliva.

Keywords: epigenetic assay, DNA methylation, PTSD, whole blood, multiplexing

Procedia PDF Downloads 124

Authors: Hugo Daneluzzo, Christelle Rabbat, Alan Jean-Marie

Abstract:

Water electrolysis is one of the most advanced technologies for producing hydrogen and can be easily combined with electricity from different sources. Under the influence of electric current, water molecules can be split into oxygen and hydrogen. The production of hydrogen by water electrolysis favors the integration of renewable energy sources into the energy mix by compensating for their intermittence through the storage of the energy produced when production exceeds demand and its release during off-peak production periods. Among the various electrolysis technologies, anion exchange membrane (AEM) electrolyser cells are emerging as a reliable technology for water electrolysis. Modeling and simulation are effective tools to save time, money, and effort during the optimization of operating conditions and the investigation of the design. The modeling and simulation become even more important when dealing with multiphysics dynamic systems. One of those systems is the AEM electrolysis cell involving complex physico-chemical reactions. Once developed, models may be utilized to comprehend the mechanisms to control and detect flaws in the systems. Several modeling methods have been initiated by scientists. These methods can be separated into two main approaches, namely equation-based modeling and graph-based modeling. The former approach is less user-friendly and difficult to update as it is based on ordinary or partial differential equations to represent the systems. However, the latter approach is more user-friendly and allows a clear representation of physical phenomena. In this case, the system is depicted by connecting subsystems, so-called blocks, through ports based on their physical interactions, hence being suitable for multiphysics systems. Among the graphical modelling methods, the bond graph is receiving increasing attention as being domain-independent and relying on the energy exchange between the components of the system. At present, few studies have investigated the modelling of AEM systems. A mathematical model and a bond graph model were used in previous studies to model the electrolysis cell performance. In this study, experimental data from literature were simulated using OpenModelica using bond graphs and mathematical approaches. The polarization curves at different operating conditions obtained by both approaches were compared with experimental ones. It was stated that both models predicted satisfactorily the polarization curves with error margins lower than 2% for equation-based models and lower than 5% for the bond graph model. The activation polarization of hydrogen evolution reactions (HER) and oxygen evolution reactions (OER) were behind the voltage loss in the AEM electrolyzer, whereas ion conduction through the membrane resulted in the ohmic loss. Therefore, highly active electro-catalysts are required for both HER and OER while high-conductivity AEMs are needed for effectively lowering the ohmic losses. The bond graph simulation of the polarisation curve for operating conditions at various temperatures has illustrated that voltage increases with temperature owing to the technology of the membrane. Simulation of the polarisation curve can be tested virtually, hence resulting in reduced cost and time involved due to experimental testing and improved design optimization. Further improvements can be made by implementing the bond graph model in a real power-to-gas-to-power scenario.

Keywords: hydrogen production, anion-exchange membrane, electrolyzer, mathematical modeling, multiphysics modeling

Procedia PDF Downloads 91

Authors: Jatziri Y. Moreno-Martinez, Arturo Galvan, Xavier Chavez Cardenas, Hiram Arroyo

Abstract:

Several methods have been utilized to study the prediction of cracking of concrete structural under loading. The finite element analysis is an alternative that shows good results. The aim of this work was the numerical study of the width crack in reinforced concrete beams with dapped ends, these are frequently found in bridge girders and precast concrete construction. Properly restricting cracking is an important aspect of the design in dapped ends, it has been observed that the cracks that exceed the allowable widths are unacceptable in an aggressive environment for reinforcing steel. For simulating the crack width, the discrete crack approach was considered by means of a Cohesive Zone (CZM) Model using a function to represent the crack opening. Two cases of dapped-end were constructed and tested in the laboratory of Structures and Materials of Engineering Institute of UNAM. The first case considers a reinforcement based on hangers as well as on vertical and horizontal ring, the second case considers 50% of the vertical stirrups in the dapped end to the main part of the beam were replaced by an equivalent area (vertically projected) of diagonal bars under. The loading protocol consisted on applying symmetrical loading to reach the service load. The models were performed using the software package ANSYS v. 16.2. The concrete structure was modeled using three-dimensional solid elements SOLID65 capable of cracking in tension and crushing in compression. Drucker-Prager yield surface was used to include the plastic deformations. The reinforcement was introduced with smeared approach. Interface delamination was modeled by traditional fracture mechanics methods such as the nodal release technique adopting softening relationships between tractions and the separations, which in turn introduce a critical fracture energy that is also the energy required to break apart the interface surfaces. This technique is called CZM. The interface surfaces of the materials are represented by a contact elements Surface-to-Surface (CONTA173) with bonded (initial contact). The Mode I dominated bilinear CZM model assumes that the separation of the material interface is dominated by the displacement jump normal to the interface. Furthermore, the opening crack was taken into consideration according to the maximum normal contact stress, the contact gap at the completion of debonding, and the maximum equivalent tangential contact stress. The contact elements were placed in the crack re-entrant corner. To validate the proposed approach, the results obtained with the previous procedure are compared with experimental test. A good correlation between the experimental and numerical Load-Displacement curves was presented, the numerical models also allowed to obtain the load-crack width curves. In these two cases, the proposed model confirms the capability of predicting the maximum crack width, with an error of ± 30 %. Finally, the orientation of the crack is a fundamental for the prediction of crack width. The results regarding the crack width can be considered as good from the practical point view. Load-Displacement curve of the test and the location of the crack were able to obtain favorable results.

Keywords: cohesive zone model, dapped-end beams, discrete crack approach, finite element analysis

Procedia PDF Downloads 167

Authors: Sofia N. Gonçalves, Duarte M. S. Albuquerque, José C. F. Pereira

Abstract:

Microwave technology has been gaining importance in contributing to decarbonization processes in high energy demand industries. Despite the several numerical models presented in the literature, a proper Verification and Validation exercise is still lacking. This is important and required to evaluate the physical process model accuracy and adequacy. Another issue addresses impedance matching, which is an important mechanism used in microwave experiments to increase electromagnetic efficiency. Such mechanism is not available in current computational tools, thus requiring an external numerical procedure. A numerical model was implemented to study the continuous processing of limestone with microwave heating. This process requires the material to be heated until a certain temperature that will prompt a highly endothermic reaction. Both a 2D and 3D model were built in COMSOL Multiphysics to solve the two-way coupling between Maxwell and Energy equations, along with the coupling between both heat transfer phenomena and limestone endothermic reaction. The 2D model was used to study and evaluate the required numerical procedure, being also a benchmark test, allowing other authors to implement impedance matching procedures. To achieve this goal, a controller built in MATLAB was used to continuously matching the cavity impedance and predicting the required energy for the system, thus successfully avoiding energy inefficiencies. The 3D model reproduces realistic results and therefore supports the main conclusions of this work. Limestone was modeled as a continuous flow under the transport of concentrated species, whose material and kinetics properties were taken from literature. Verification and Validation of the coupled model was taken separately from the chemical kinetic model. The chemical kinetic model was found to correctly describe the chosen kinetic equation by comparing numerical results with experimental data. A solution verification was made for the electromagnetic interface, where second order and fourth order accurate schemes were found for linear and quadratic elements, respectively, with numerical uncertainty lower than 0.03%. Regarding the coupled model, it was demonstrated that the numerical error would diverge for the heat transfer interface with the mapped mesh. Results showed numerical stability for the triangular mesh, and the numerical uncertainty was less than 0.1%. This study evaluated limestone velocity, heat transfer, and load influence on thermal decomposition and overall process efficiency. The velocity and heat transfer coefficient were studied with the 2D model, while different loads of material were studied with the 3D model. Both models demonstrated to be highly unstable when solving non-linear temperature distributions. High velocity flows exhibited propensity to thermal runways, and the thermal efficiency showed the tendency to stabilize for the higher velocities and higher filling ratio. Microwave efficiency denoted an optimal velocity for each heat transfer coefficient, pointing out that electromagnetic efficiency is a consequence of energy distribution uniformity. The 3D results indicated the inefficient development of the electric field for low filling ratios. Thermal efficiencies higher than 90% were found for the higher loads and microwave efficiencies up to 75% were accomplished. The 80% fill ratio was demonstrated to be the optimal load with an associated global efficiency of 70%.

Keywords: multiphysics modeling, microwave heating, verification and validation, endothermic reactions modeling, impedance matching, limestone continuous processing

Procedia PDF Downloads 140

Authors: Tolulope Aremu

Abstract:

The key process steps to produce liquid detergent products will introduce potential defects, such as formulation, mixing, filling, and packaging, which might compromise product quality, consumer safety, and operational efficiency. Real-time identification and characterization of such defects are of prime importance for maintaining high standards and reducing waste and costs. Usually, defect detection is performed by human inspection or rule-based systems, which is very time-consuming, inconsistent, and error-prone. The present study overcomes these limitations in dealing with optimization in defect characterization within the process for making liquid detergents using Machine Learning algorithms. Performance testing of various machine learning models was carried out: Support Vector Machine, Decision Trees, Random Forest, and Convolutional Neural Network on defect detection and classification of those defects like wrong viscosity, color deviations, improper filling of a bottle, packaging anomalies. These algorithms have significantly benefited from a variety of optimization techniques, including hyperparameter tuning and ensemble learning, in order to greatly improve detection accuracy while minimizing false positives. Equipped with a rich dataset of defect types and production parameters consisting of more than 100,000 samples, our study further includes information from real-time sensor data, imaging technologies, and historic production records. The results are that optimized machine learning models significantly improve defect detection compared to traditional methods. Take, for instance, the CNNs, which run at 98% and 96% accuracy in detecting packaging anomaly detection and bottle filling inconsistency, respectively, by fine-tuning the model with real-time imaging data, through which there was a reduction in false positives of about 30%. The optimized SVM model on detecting formulation defects gave 94% in viscosity variation detection and color variation. These values of performance metrics correspond to a giant leap in defect detection accuracy compared to the usual 80% level achieved up to now by rule-based systems. Moreover, this optimization with models can hasten defect characterization, allowing for detection time to be below 15 seconds from an average of 3 minutes using manual inspections with real-time processing of data. With this, the reduction in time will be combined with a 25% reduction in production downtime because of proactive defect identification, which can save millions annually in recall and rework costs. Integrating real-time machine learning-driven monitoring drives predictive maintenance and corrective measures for a 20% improvement in overall production efficiency. Therefore, the optimization of machine learning algorithms in defect characterization optimum scalability and efficiency for liquid detergent companies gives improved operational performance to higher levels of product quality. In general, this method could be conducted in several industries within the Fast moving consumer Goods industry, which would lead to an improved quality control process.

Keywords: liquid detergent manufacturing, defect detection, machine learning, support vector machines, convolutional neural networks, defect characterization, predictive maintenance, quality control, fast-moving consumer goods

Procedia PDF Downloads 19

Authors: Chung-Ming Lo, Chung-Chien Lee

Abstract:

In many countries, the lifetime prevalence of shoulder pain is up to 70%. In America, the health care system spends 7 billion per year about the healthy issues of shoulder pain. With respect to the origin, up to 70% of shoulder pain is attributed to rotator cuff lesions This study proposed a computer-aided diagnosis (CAD) system to assist radiologists classifying rotator cuff lesions with less operator dependence. Quantitative features were extracted from the shoulder ultrasound images acquired using an ALOKA alpha-6 US scanner (Hitachi-Aloka Medical, Tokyo, Japan) with linear array probe (scan width: 36mm) ranging from 5 to 13 MHz. During examination, the postures of the examined patients are standard sitting position and are followed by the regular routine. After acquisition, the shoulder US images were drawn out from the scanner and stored as 8-bit images with pixel value ranging from 0 to 255. Upon the sonographic appearance, the boundary of each lesion was delineated by a physician to indicate the specific pattern for analysis. The three lesion categories for classification were composed of 20 cases of tendon inflammation, 18 cases of calcific tendonitis, and 18 cases of supraspinatus tear. For each lesion, second-order statistics were quantified in the feature extraction. The second-order statistics were the texture features describing the correlations between adjacent pixels in a lesion. Because echogenicity patterns were expressed via grey-scale. The grey-scale co-occurrence matrixes with four angles of adjacent pixels were used. The texture metrics included the mean and standard deviation of energy, entropy, correlation, inverse different moment, inertia, cluster shade, cluster prominence, and Haralick correlation. Then, the quantitative features were combined in a multinomial logistic regression classifier to generate a prediction model of rotator cuff lesions. Multinomial logistic regression classifier is widely used in the classification of more than two categories such as the three lesion types used in this study. In the classifier, backward elimination was used to select a feature subset which is the most relevant. They were selected from the trained classifier with the lowest error rate. Leave-one-out cross-validation was used to evaluate the performance of the classifier. Each case was left out of the total cases and used to test the trained result by the remaining cases. According to the physician’s assessment, the performance of the proposed CAD system was shown by the accuracy. As a result, the proposed system achieved an accuracy of 86%. A CAD system based on the statistical texture features to interpret echogenicity values in shoulder musculoskeletal ultrasound was established to generate a prediction model for rotator cuff lesions. Clinically, it is difficult to distinguish some kinds of rotator cuff lesions, especially partial-thickness tear of rotator cuff. The shoulder orthopaedic surgeon and musculoskeletal radiologist reported greater diagnostic test accuracy than general radiologist or ultrasonographers based on the available literature. Consequently, the proposed CAD system which was developed according to the experiment of the shoulder orthopaedic surgeon can provide reliable suggestions to general radiologists or ultrasonographers. More quantitative features related to the specific patterns of different lesion types would be investigated in the further study to improve the prediction.

Keywords: shoulder ultrasound, rotator cuff lesions, texture, computer-aided diagnosis

Procedia PDF Downloads 284

Authors: Rahil Bahrami, Kaveh Ashenayi

Abstract:

This paper presents a systematic classification and comparative analysis of non-isolated bi-directional DC-DC converters. The increasing demand for efficient energy conversion in diverse applications has spurred the development of various converter topologies. In this study, we categorize bi-directional converters into three distinct classes: Inverting, Non-Inverting, and Interleaved. Each category is characterized by its unique operational characteristics and benefits. Furthermore, a practical comparison is conducted by evaluating the results of simulation of each bi-directional converter. BDCs can be classified into isolated and non-isolated topologies. Non-isolated converters share a common ground between input and output, making them suitable for applications with minimal voltage change. They are easy to integrate, lightweight, and cost-effective but have limitations like limited voltage gain, switching losses, and no protection against high voltages. Isolated converters use transformers to separate input and output, offering safety benefits, high voltage gain, and noise reduction. They are larger and more costly but are essential for automotive designs where safety is crucial. The paper focuses on non-isolated systems.The paper discusses the classification of non-isolated bidirectional converters based on several criteria. Common factors used for classification include topology, voltage conversion, control strategy, power capacity, voltage range, and application. These factors serve as a foundation for categorizing converters, although the specific scheme might vary depending on contextual, application, or system-specific requirements. The paper presents a three-category classification for non-isolated bi-directional DC-DC converters: inverting, non-inverting, and interleaved. In the inverting category, converters produce an output voltage with reversed polarity compared to the input voltage, achieved through specific circuit configurations and control strategies. This is valuable in applications such as motor control and grid-tied solar systems. The non-inverting category consists of converters maintaining the same voltage polarity, useful in scenarios like battery equalization. Lastly, the interleaved category employs parallel converter stages to enhance power delivery and reduce current ripple. This classification framework enhances comprehension and analysis of non-isolated bi-directional DC-DC converters. The findings contribute to a deeper understanding of the trade-offs and merits associated with different converter types. As a result, this work aids researchers, practitioners, and engineers in selecting appropriate bi-directional converter solutions for specific energy conversion requirements. The proposed classification framework and experimental assessment collectively enhance the comprehension of non-isolated bi-directional DC-DC converters, fostering advancements in efficient power management and utilization.The simulation process involves the utilization of PSIM to model and simulate non-isolated bi-directional converter from both inverted and non-inverted category. The aim is to conduct a comprehensive comparative analysis of these converters, considering key performance indicators such as rise time, efficiency, ripple factor, and maximum error. This systematic evaluation provides valuable insights into the dynamic response, energy efficiency, output stability, and overall precision of the converters. The results of this comparison facilitate informed decision-making and potential optimizations, ensuring that the chosen converter configuration aligns effectively with the designated operational criteria and performance goals.

Keywords: bi-directional, DC-DC converter, non-isolated, energy conversion

Procedia PDF Downloads 100

Authors: Przemysław Adamczyk, Mirosław Wyczesany, Aleksandra Domagalik, Artur Daren, Kamil Cepuch, Piotr Błądziński, Tadeusz Marek, Andrzej Cechnicki

Abstract:

The present study aimed at evaluation of neural correlates of humor comprehension impairments observed in schizophrenia. To investigate the nature of this deficit in schizophrenia and to localize cortical areas involved in humor processing we used functional magnetic resonance imaging (fMRI). The study included chronic schizophrenia outpatients (SCH; n=20), and sex, age and education level matched healthy controls (n=20). The task consisted of 60 stories (setup) of which 20 had funny, 20 nonsensical and 20 neutral (not funny) punchlines. After the punchlines were presented, the participants were asked to indicate whether the story was comprehensible (yes/no) and how funny it was (1-9 Likert-type scale). fMRI was performed on a 3T scanner (Magnetom Skyra, Siemens) using 32-channel head coil. Three contrasts in accordance with the three stages of humor processing were analyzed in both groups: abstract vs neutral stories - incongruity detection; funny vs abstract - incongruity resolution; funny vs neutral - elaboration. Additionally, parametric modulation analysis was performed using both subjective ratings separately in order to further differentiate the areas involved in incongruity resolution processing. Statistical analysis for behavioral data used U Mann-Whitney test and Bonferroni’s correction, fMRI data analysis utilized whole-brain voxel-wise t-tests with 10-voxel extent threshold and with Family Wise Error (FWE) correction at alpha = 0.05, or uncorrected at alpha = 0.001. Between group comparisons revealed that the SCH subjects had attenuated activation in: the right superior temporal gyrus in case of irresolvable incongruity processing of nonsensical puns (nonsensical > neutral); the left medial frontal gyrus in case of incongruity resolution processing of funny puns (funny > nonsensical) and the interhemispheric ACC in case of elaboration of funny puns (funny > neutral). Additionally, the SCH group revealed weaker activation during funniness ratings in the left ventro-medial prefrontal cortex, the medial frontal gyrus, the angular and the supramarginal gyrus, and the right temporal pole. In comprehension ratings the SCH group showed suppressed activity in the left superior and medial frontal gyri. Interestingly, these differences were accompanied by protraction of time in both types of rating responses in the SCH group, a lower level of comprehension for funny punchlines and a higher funniness for absurd punchlines. Presented results indicate that, in comparison to healthy controls, schizophrenia is characterized by difficulties in humor processing revealed by longer reaction times, impairments of understanding jokes and finding nonsensical punchlines more funny. This is accompanied by attenuated brain activations, especially in the left fronto-parietal and the right temporal cortices. Disturbances of the humor processing seem to be impaired at the all three stages of the humor comprehension process, from incongruity detection, through its resolution to elaboration. The neural correlates revealed diminished neural activity of the schizophrenia brain, as compared with the control group. The study was supported by the National Science Centre, Poland (grant no 2014/13/B/HS6/03091).

Keywords: communication skills, functional magnetic resonance imaging, humor, schizophrenia

Procedia PDF Downloads 213

Authors: Claire Williams

Abstract:

Randomised controlled Trials (RCTs) of efficacy are still perceived as the gold standard for the generation of evidence, and whilst advances in data collection methods are well developed, this progress has not been matched for the reporting of adverse events (AEs). Assessment and reporting of AEs in clinical trials are fraught with human error and inefficiency and are extremely time and resource intensive. Recent research conducted into the quality of reporting of AEs during clinical trials concluded it is substandard and reporting is inconsistent. Investigators commonly send reports to sponsors who are incorrectly categorised and lacking in critical information, which can complicate the detection of valid safety signals. In our presentation, we will describe an electronic data capture system, which has been designed to support clinical trial processes by reducing the resource burden on investigators, improving overall trial efficiencies, and making trials safer for patients. This proprietary technology was developed using expertise proven in the delivery of the world’s first prospective, phase 3b real-world trial, ‘The Salford Lung Study, ’ which enabled robust safety monitoring and reporting processes to be accomplished by the remote monitoring of patients’ EHRs. This technology enables safety alerts that are pre-defined by the protocol to be detected from the data extracted directly from the patients EHR. Based on study-specific criteria, which are created from the standard definition of a serious adverse event (SAE) and the safety profile of the medicinal product, the system alerts the investigator or study team to the safety alert. Each safety alert will require a clinical review by the investigator or delegate; examples of the types of alerts include hospital admission, death, hepatotoxicity, neutropenia, and acute renal failure. This is achieved in near real-time; safety alerts can be reviewed along with any additional information available to determine whether they meet the protocol-defined criteria for reporting or withdrawal. This active surveillance technology helps reduce the resource burden of the more traditional methods of AE detection for the investigators and study teams and can help eliminate reporting bias. Integration of multiple healthcare data sources enables much more complete and accurate safety data to be collected as part of a trial and can also provide an opportunity to evaluate a drug’s safety profile long-term, in post-trial follow-up. By utilising this robust and proven method for safety monitoring and reporting, a much higher risk of patient cohorts can be enrolled into trials, thus promoting inclusivity and diversity. Broadening eligibility criteria and adopting more inclusive recruitment practices in the later stages of drug development will increase the ability to understand the medicinal products risk-benefit profile across the patient population that is likely to use the product in clinical practice. Furthermore, this ground-breaking approach to AE detection not only provides sponsors with better-quality safety data for their products, but it reduces the resource burden on the investigator and study teams. With the data taken directly from the source, trial costs are reduced, with minimal data validation required and near real-time reporting enables safety concerns and signals to be detected more quickly than in a traditional RCT.

Keywords: more comprehensive and accurate safety data, near real-time safety alerts, reduced resource burden, safer trials

Procedia PDF Downloads 85

Authors: G. K. Viswanadh, B. Rajasekhar, G. Venkata Ramana

Abstract:

Water distribution networks play a vital role in ensuring a reliable supply of clean water to urban areas. However, they face several challenges, including pressure control, pump speed optimization, and burst event detection. This paper combines insights from two studies to address these critical issues in Water distribution networks, focusing on the specific context of Kapra Municipality, India. The first part of this research concentrates on optimizing pressure control and pump speed in complex Water distribution networks. It utilizes the EPANET- MATLAB Toolkit to integrate EPANET functionalities into the MATLAB environment, offering a comprehensive approach to network analysis. By optimizing Pressure Reduce Valves (PRVs) and variable speed pumps (VSPs), this study achieves remarkable results. In the Benchmark Water Distribution System (WDS), the proposed PRV optimization algorithm reduces average leakage by 20.64%, surpassing the previous achievement of 16.07%. When applied to the South-Central and East zone WDS of Kapra Municipality, it identifies PRV locations that were previously missed by existing algorithms, resulting in average leakage reductions of 22.04% and 10.47%. These reductions translate to significant daily Water savings, enhancing Water supply reliability and reducing energy consumption. The second part of this research addresses the pressing issue of burst event detection and localization within the Water Distribution System. Burst events are a major contributor to Water losses and repair expenses. The study employs wireless sensor technology to monitor pressure and flow rate in real time, enabling the detection of pipeline abnormalities, particularly burst events. The methodology relies on transient analysis of pressure signals, utilizing Cumulative Sum and Wavelet analysis techniques to robustly identify burst occurrences. To enhance precision, burst event localization is achieved through meticulous analysis of time differentials in the arrival of negative pressure waveforms across distinct pressure sensing points, aided by nodal matrix analysis. To evaluate the effectiveness of this methodology, a PVC Water pipeline test bed is employed, demonstrating the algorithm's success in detecting pipeline burst events at flow rates of 2-3 l/s. Remarkably, the algorithm achieves a localization error of merely 3 meters, outperforming previously established algorithms. This research presents a significant advancement in efficient burst event detection and localization within Water pipelines, holding the potential to markedly curtail Water losses and the concomitant financial implications. In conclusion, this combined research addresses critical challenges in Water distribution networks, offering solutions for optimizing pressure control, pump speed, burst event detection, and localization. These findings contribute to the enhancement of Water Distribution System, resulting in improved Water supply reliability, reduced Water losses, and substantial cost savings. The integrated approach presented in this paper holds promise for municipalities and utilities seeking to improve the efficiency and sustainability of their Water distribution networks.

Keywords: pressure reduce valve, complex networks, variable speed pump, wavelet transform, burst detection, CUSUM (Cumulative Sum), water pipeline monitoring

Procedia PDF Downloads 87

Authors: Ho-Ting Wu, Kai-Wei Ke, Bo-Yu Huang, Liang-Lin Yan, Chun-Ting Lin

Abstract:

With the advances in advanced technology, wireless sensor network (WSN) has become one of the most promising candidates to implement the wireless industrial internet of things (IIOT) architecture. However, the legacy IEEE 802.15.4 based WSN technology such as Zigbee system cannot meet the stringent QoS requirement of low powered, real-time, and highly reliable transmission imposed by the IIOT environment. Recently, the IEEE society developed IEEE 802.15.4e Time Slotted Channel Hopping (TSCH) access mode to serve this purpose. Furthermore, the IETF 6TiSCH working group has proposed standards to integrate IEEE 802.15.4e with IPv6 protocol smoothly to form a complete protocol stack for IIOT. In this work, we develop key network technologies for IEEE 802.15.4e based wireless IIoT architecture, focusing on practical design and system implementation. We realize the OpenWSN-based wireless IIOT system. The system architecture is divided into three main parts: web server, network manager, and sensor nodes. The web server provides user interface, allowing the user to view the status of sensor nodes and instruct sensor nodes to follow commands via user-friendly browser. The network manager is responsible for the establishment, maintenance, and management of scheduling and topology information. It executes centralized scheduling algorithm, sends the scheduling table to each node, as well as manages the sensing tasks of each device. Sensor nodes complete the assigned tasks and sends the sensed data. Furthermore, to prevent scheduling error due to packet loss, a schedule inspection mechanism is implemented to verify the correctness of the schedule table. In addition, when network topology changes, the system will act to generate a new schedule table based on the changed topology for ensuring the proper operation of the system. To enhance the system performance of such system, we further propose dynamic bandwidth allocation and distributed scheduling mechanisms. The developed distributed scheduling mechanism enables each individual sensor node to build, maintain and manage the dedicated link bandwidth with its parent and children nodes based on locally observed information by exchanging the Add/Delete commands via two processes. The first process, termed as the schedule initialization process, allows each sensor node pair to identify the available idle slots to allocate the basic dedicated transmission bandwidth. The second process, termed as the schedule adjustment process, enables each sensor node pair to adjust their allocated bandwidth dynamically according to the measured traffic loading. Such technology can sufficiently satisfy the dynamic bandwidth requirement in the frequently changing environments. Last but not least, we propose a packet retransmission scheme to enhance the system performance of the centralized scheduling algorithm when the packet delivery rate (PDR) is low. We propose a multi-frame retransmission mechanism to allow every single network node to resend each packet for at least the predefined number of times. The multi frame architecture is built according to the number of layers of the network topology. Performance results via simulation reveal that such retransmission scheme is able to provide sufficient high transmission reliability while maintaining low packet transmission latency. Therefore, the QoS requirement of IIoT can be achieved.

Keywords: IEEE 802.15.4e, industrial internet of things (IIOT), scheduling mechanisms, wireless sensor networks (WSN)

Procedia PDF Downloads 160

Authors: Dongping Fang, Lian Duan, Xiaojing Yuan, Mike Xu, Allyn Klunder, Kevin Tan, Suiting Cao, Yeqing Ji

Abstract:

Accurate prediction of a medical condition with straight clinical evidence is a long-sought topic in the medical management and health insurance field. Although great progress has been made with machine learning algorithms, the medical community is still, to a certain degree, suspicious about the model's accuracy and interpretability. This paper presents an innovative hierarchical attention deep learning model to achieve good prediction and clear interpretability that can be easily understood by medical professionals. This deep learning model uses a hierarchical attention structure that matches naturally with the medical history data structure and reflects the member’s encounter (date of service) sequence. The model attention structure consists of 3 levels: (1) attention on the medical code types (diagnosis codes, procedure codes, lab test results, and prescription drugs), (2) attention on the sequential medical encounters within a type, (3) attention on the medical codes within an encounter and type. This model is applied to predict the occurrence of stage 3 chronic kidney disease (CKD3), using three years’ medical history of Medicare Advantage (MA) members from a top health insurance company. The model takes members’ medical events, both claims and electronic medical record (EMR) data, as input, makes a prediction of CKD3 and calculates the contribution from individual events to the predicted outcome. The model outcome can be easily explained with the clinical evidence identified by the model algorithm. Here are examples: Member A had 36 medical encounters in the past three years: multiple office visits, lab tests and medications. The model predicts member A has a high risk of CKD3 with the following well-contributed clinical events - multiple high ‘Creatinine in Serum or Plasma’ tests and multiple low kidneys functioning ‘Glomerular filtration rate’ tests. Among the abnormal lab tests, more recent results contributed more to the prediction. The model also indicates regular office visits, no abnormal findings of medical examinations, and taking proper medications decreased the CKD3 risk. Member B had 104 medical encounters in the past 3 years and was predicted to have a low risk of CKD3, because the model didn’t identify diagnoses, procedures, or medications related to kidney disease, and many lab test results, including ‘Glomerular filtration rate’ were within the normal range. The model accurately predicts members A and B and provides interpretable clinical evidence that is validated by clinicians. Without extra effort, the interpretation is generated directly from the model and presented together with the occurrence date. Our model uses the medical data in its most raw format without any further data aggregation, transformation, or mapping. This greatly simplifies the data preparation process, mitigates the chance for error and eliminates post-modeling work needed for traditional model explanation. To our knowledge, this is the first paper on an interpretable deep-learning model using a 3-level attention structure, sourcing both EMR and claim data, including all 4 types of medical data, on the entire Medicare population of a big insurance company, and more importantly, directly generating model interpretation to support user decision. In the future, we plan to enrich the model input by adding patients’ demographics and information from free-texted physician notes.

Keywords: deep learning, interpretability, attention, big data, medical conditions

Procedia PDF Downloads 91

Authors: Asma Omri, Iheb Sifaoui, Sofiane Sayahi, Hichem Besbes

Abstract:

Future vehicle systems demand advanced capabilities, notably in-cabin life detection and driver monitoring systems, with a particular emphasis on drowsiness detection. To meet these requirements, several techniques employ artificial intelligence methods based on real-time vital sign measurements. In parallel, Frequency-Modulated Continuous-Wave (FMCW) radar technology has garnered considerable attention in the domains of healthcare and biomedical engineering for non-invasive vital sign monitoring. FMCW radar offers a multitude of advantages, including its non-intrusive nature, continuous monitoring capacity, and its ability to penetrate through clothing. In this paper, we propose a system utilizing the AWR6843AOP radar from Texas Instruments (TI) to extract precise vital sign information. The radar allows us to estimate Ballistocardiogram (BCG) signals, which capture the mechanical movements of the body, particularly the ballistic forces generated by heartbeats and respiration. These signals are rich sources of information about the cardiac cycle, rendering them suitable for heart rate estimation. The process begins with real-time subject positioning, followed by clutter removal, computation of Doppler phase differences, and the use of various filtering methods to accurately capture subtle physiological movements. To address the challenges associated with FMCW radar-based vital sign monitoring, including motion artifacts due to subjects' movement or radar micro-vibrations, Long Short-Term Memory (LSTM) networks are implemented. LSTM's adaptability to different heart rate patterns and ability to handle real-time data make it suitable for continuous monitoring applications. Several crucial steps were taken, including feature extraction (involving amplitude, time intervals, and signal morphology), sequence modeling, heart rate estimation through the analysis of detected cardiac cycles and their temporal relationships, and performance evaluation using metrics such as Root Mean Square Error (RMSE) and correlation with reference heart rate measurements. For dataset construction and LSTM training, a comprehensive data collection system was established, integrating the AWR6843AOP radar, a Heart Rate Belt, and a smart watch for ground truth measurements. Rigorous synchronization of these devices ensured data accuracy. Twenty participants engaged in various scenarios, encompassing indoor and real-world conditions within a moving vehicle equipped with the radar system. Static and dynamic subject’s conditions were considered. The heart rate estimation through LSTM outperforms traditional signal processing techniques that rely on filtering, Fast Fourier Transform (FFT), and thresholding. It delivers an average accuracy of approximately 91% with an RMSE of 1.01 beat per minute (bpm). In conclusion, this paper underscores the promising potential of FMCW radar technology integrated with artificial intelligence algorithms in the context of automotive applications. This innovation not only enhances road safety but also paves the way for its integration into the automotive ecosystem to improve driver well-being and overall vehicular safety.

Keywords: ballistocardiogram, FMCW Radar, vital sign monitoring, LSTM

Procedia PDF Downloads 72

Authors: Mian Huang, Chi Ma, Junyu Lin, William Lu

Abstract:

Introduction: Bone health attracts more attention recently and an intelligent question and answer (QA) chatbot for osteoporosis is helpful for science popularization. With Generative Pretrained Transformer (GPT) technology developing, we build an osteoporosis corpus dataset and then fine-tune LLaMA, a famous open-source GPT foundation large language model(LLM), on our self-constructed osteoporosis corpus. Evaluated by clinical orthopedic experts, our fine-tuned model outperforms vanilla LLaMA on osteoporosis QA task in Chinese. Three-dimensional quantitative computed tomography (QCT) measured bone mineral density (BMD) is considered as more accurate than DXA for BMD measurement in recent years. We develop an automatic Phantom-less QCT(PL-QCT) that is more efficient for BMD measurement since no need of an external phantom for calibration. Combined with LLM on osteoporosis, our PL-QCT provides efficient and accurate BMD measurement for our chatbot users. Material and Methods: We build an osteoporosis corpus containing about 30,000 Chinese literatures whose titles are related to osteoporosis. The whole process is done automatically, including crawling literatures in .pdf format, localizing text/figure/table region by layout segmentation algorithm and recognizing text by OCR algorithm. We train our model by continuous pre-training with Low-rank Adaptation (LoRA, rank=10) technology to adapt LLaMA-7B model to osteoporosis domain, whose basic principle is to mask the next word in the text and make the model predict that word. The loss function is defined as cross-entropy between the predicted and ground-truth word. Experiment is implemented on single NVIDIA A800 GPU for 15 days. Our automatic PL-QCT BMD measurement adopt AI-associated region-of-interest (ROI) generation algorithm for localizing vertebrae-parallel cylinder in cancellous bone. Due to no phantom for BMD calibration, we calculate ROI BMD by CT-BMD of personal muscle and fat. Results & Discussion: Clinical orthopaedic experts are invited to design 5 osteoporosis questions in Chinese, evaluating performance of vanilla LLaMA and our fine-tuned model. Our model outperforms LLaMA on over 80% of these questions, understanding ‘Expert Consensus on Osteoporosis’, ‘QCT for osteoporosis diagnosis’ and ‘Effect of age on osteoporosis’. Detailed results are shown in appendix. Future work may be done by training a larger LLM on the whole orthopaedics with more high-quality domain data, or a multi-modal GPT combining and understanding X-ray and medical text for orthopaedic computer-aided-diagnosis. However, GPT model gives unexpected outputs sometimes, such as repetitive text or seemingly normal but wrong answer (called ‘hallucination’). Even though GPT give correct answers, it cannot be considered as valid clinical diagnoses instead of clinical doctors. The PL-QCT BMD system provided by Bone’s QCT(Bone’s Technology(Shenzhen) Limited) achieves 0.1448mg/cm2(spine) and 0.0002 mg/cm2(hip) mean absolute error(MAE) and linear correlation coefficient R2=0.9970(spine) and R2=0.9991(hip)(compared to QCT-Pro(Mindways)) on 155 patients in three-center clinical trial in Guangzhou, China. Conclusion: This study builds a Chinese osteoporosis corpus and develops a fine-tuned and domain-adapted LLM as well as a PL-QCT BMD measurement system. Our fine-tuned GPT model shows better capability than LLaMA model on most testing questions on osteoporosis. Combined with our PL-QCT BMD system, we are looking forward to providing science popularization and early morning screening for potential osteoporotic patients.

Keywords: GPT, phantom-less QCT, large language model, osteoporosis

Procedia PDF Downloads 71