Search results for: neural electrodes

Authors: Samuel J. Ahmad, Michael Zhu, Andrew J. Kobets

Abstract:

Segmentation tools continue to advance, evolving from manual methods to automated contouring technologies utilizing convolutional neural networks. These techniques have evaluated ventricular and hemorrhagic volumes in the past but may be applied in novel ways to assess posterior fossa-associated pathologies such as Chiari malformations. Herein, we summarize literature pertaining to segmentation in the context of this and other posterior fossa-based diseases such as trigeminal neuralgia, hemifacial spasm, and posterior fossa syndrome. A literature search for volumetric analysis of the posterior fossa identified 27 papers where semi-automated, automated, manual segmentation, linear measurement-based formulas, and the Cavalieri estimator were utilized. These studies produced superior data than older methods utilizing formulas for rough volumetric estimations. The most commonly used segmentation technique was semi-automated segmentation (12 studies). Manual segmentation was the second most common technique (7 studies). Automated segmentation techniques (4 studies) and the Cavalieri estimator (3 studies), a point-counting method that uses a grid of points to estimate the volume of a region, were the next most commonly used techniques. The least commonly utilized segmentation technique was linear measurement-based formulas (1 study). Semi-automated segmentation produced accurate, reproducible results. However, it is apparent that there does not exist a single semi-automated software, open source or otherwise, that has been widely applied to the posterior fossa. Fully-automated segmentation via such open source software as FSL and Freesurfer produced highly accurate posterior fossa segmentations. Various forms of segmentation have been used to assess posterior fossa pathologies and each has its advantages and disadvantages. According to our results, semi-automated segmentation is the predominant method. However, atlas-based automated segmentation is an extremely promising method that produces accurate results. Future evolution of segmentation technologies will undoubtedly yield superior results, which may be applied to posterior fossa related pathologies. Medical professionals will save time and effort analyzing large sets of data due to these advances.

Keywords: chiari, posterior fossa, segmentation, volumetric

Procedia PDF Downloads 107

Authors: Breanna Macumber, Victor A. Huayamave, Emir A. Vela, Wangdo Kim, Tamara T. Chamber, Esteban Centeno

Abstract:

Spina bifida is a type of neural tube defect that affects the nervous system and can lead to problems such as total leg paralysis. Treatment requires physical therapy and rehabilitation. Robotic exoskeletons have been used for rehabilitation to train muscle movement and assist in injury recovery; however, current models focus on the adult populations and not on the infant population. The proposed framework aims to couple a musculoskeletal infant model with a robotic exoskeleton using vacuum-powered artificial muscles to provide rehabilitation to infants affected by spina bifida. The study that drove the input values for the robotic exoskeleton used motion capture technology to collect data from the spontaneous kicking movement of a 2.4-month-old infant lying supine. OpenSim was used to develop the musculoskeletal model, and Inverse kinematics was used to estimate hip joint angles. A total of 4 kicks (A, B, C, D) were selected, and the selection was based on range, transient response, and stable response. Kicks had at least 5° of range of motion with a smooth transient response and a stable period. The robotic exoskeleton used a Vacuum-Powered Artificial Muscle (VPAM) the structure comprised of cells that were clipped in a collapsed state and unclipped when desired to simulate infant’s age. The artificial muscle works with vacuum pressure. When air is removed, the muscle contracts and when air is added, the muscle relaxes. Bench testing was performed using a 6-month-old infant mannequin. The previously developed exoskeleton worked really well with controlled ranges of motion and frequencies, which are typical of rehabilitation protocols for infants suffering with spina bifida. However, the random kicking motion in this study contained high frequency kicks and was not able to accurately replicate all the investigated kicks. Kick 'A' had a greater error when compared to the other kicks. This study has the potential to advance the infant rehabilitation field.

Keywords: musculoskeletal modeling, soft robotics, rehabilitation, pediatrics

Procedia PDF Downloads 120

Authors: Pejman Salimi, Sebastiano Tieuli, Somayeh Taghavi, Michela Signoretto, Remo Proietti Zaccaria

Abstract:

Increasing the growth of industrial waste due to the large quantities of production leads to numerous environmental and economic challenges, such as climate change, soil and water contamination, human disease, etc. Energy recovery of waste can be applied to produce heat or electricity. This strategy allows for the reduction of energy produced using coal or other fuels and directly reduces greenhouse gas emissions. Among different factories, leather manufacturing plays a very important role in the whole world from the socio-economic point of view. The leather industry plays a very important role in our society from a socio-economic point of view. Even though the leather industry uses a by-product from the meat industry as raw material, it is considered as an activity demanding integrated prevention and control of pollution. Along the entire process from raw skins/hides to finished leather, a huge amount of solid and water waste is generated. Solid wastes include fleshings, raw trimmings, shavings, buffing dust, etc. One of the most abundant solid wastes generated throughout leather tanning is shaving waste. Leather shaving is a mechanical process that aims at reducing the tanned skin to a specific thickness before tanning and finishing. This product consists mainly of collagen and tanning agent. At present, most of the world's leather processing is chrome-tanned based. Consequently, large amounts of chromium-containing shaving wastes need to be treated. The major concern about the management of this kind of solid waste is ascribed to chrome content, which makes the conventional disposal methods, such as landfilling and incineration, not practicable. Therefore, many efforts have been developed in recent decades to promote eco-friendly/alternative leather production and more effective waste management. Herein, shaving waste resulting from metal-free tanning technology is proposed as low-cost precursors for the preparation of carbon material as anodes for lithium-ion batteries (LIBs). In line with the philosophy of a reduced environmental impact, for preparing fully sustainable and environmentally friendly LIBs anodes, deionized water and carboxymethyl cellulose (CMC) have been used as alternatives to toxic/teratogen N-methyl-2- pyrrolidone (NMP) and to biologically hazardous Polyvinylidene fluoride (PVdF), respectively. Furthermore, going towards the reduced cost, we employed water solvent and fluoride-free bio-derived CMC binder (as an alternative to NMP and PVdF, respectively) together with LiFePO₄ (LFP) when a full cell was considered. These actions make closer to the 2030 goal of having green LIBs at 100 $ kW h⁻¹. Besides, the preparation of the water-based electrodes does not need a controlled environment and due to the higher vapour pressure of water in comparison with NMP, the water-based electrode drying is much faster. This aspect determines an important consequence, namely a reduced energy consumption for the electrode preparation. The electrode derived from leather waste demonstrated a discharge capacity of 735 mAh g⁻¹ after 1000 charge and discharge cycles at 0.5 A g⁻¹. This promising performance is ascribed to the synergistic effect of defects, interlayer spacing, heteroatoms-doped (N, O, and S), high specific surface area, and hierarchical micro/mesopore structure of the biochar. Interestingly, these features of activated biochars derived from the leather industry open the way for possible applications in other EESDs as well.

Keywords: biowaste, lithium-ion batteries, physical activation, waste management, leather industry

Procedia PDF Downloads 171

Authors: Breanna Macumber, Victor A. Huayamave, Emir A. Vela, Wangdo Kim, Tamara T. Chamber, Esteban Centeno

Abstract:

Spina bifida is a type of neural tube defect that affects the nervous system and can lead to problems such as total leg paralysis. Treatment requires physical therapy and rehabilitation. Robotic exoskeletons have been used for rehabilitation to train muscle movement and assist in injury recovery; however, current models focus on the adult populations and not on the infant population. The proposed framework aims to couple a musculoskeletal infant model with a robotic exoskeleton using vacuum-powered artificial muscles to provide rehabilitation to infants affected by spina bifida. The study that drove the input values for the robotic exoskeleton used motion capture technology to collect data from the spontaneous kicking movement of a 2.4-month-old infant lying supine. OpenSim was used to develop the musculoskeletal model, and Inverse kinematics was used to estimate hip joint angles. A total of 4 kicks (A, B, C, D) were selected, and the selection was based on range, transient response, and stable response. Kicks had at least 5° of range of motion with a smooth transient response and a stable period. The robotic exoskeleton used a Vacuum-Powered Artificial Muscle (VPAM) the structure comprised of cells that were clipped in a collapsed state and unclipped when desired to simulate infant’s age. The artificial muscle works with vacuum pressure. When air is removed, the muscle contracts and when air is added, the muscle relaxes. Bench testing was performed using a 6-month-old infant mannequin. The previously developed exoskeleton worked really well with controlled ranges of motion and frequencies, which are typical of rehabilitation protocols for infants suffering with spina bifida. However, the random kicking motion in this study contained high frequency kicks and was not able to accurately replicate all the investigated kicks. Kick 'A' had a greater error when compared to the other kicks. This study has the potential to advance the infant rehabilitation field.

Keywords: musculoskeletal modeling, soft robotics, rehabilitation, pediatrics

Procedia PDF Downloads 88

Authors: Ferran Paulí Pla, Pere Palacín Farré, Albert Fornells Herrera, Pol Toldrà Fernández

Abstract:

Poor image acquisition is limiting the promising growth of industrial vision in food control. In recent years, the food industry has witnessed a significant increase in the implementation of automation in quality control through artificial vision, a trend that continues to grow. During the packaging process, some defects may appear, compromising the proper sealing of the products and diminishing their shelf life, sanitary conditions and overall properties. While failure to detect a defective product leads to major losses, food producers also aim to minimize over-rejection to avoid unnecessary waste. Thus, accuracy in the evaluation of the products is crucial, and, given the large production volumes, even small improvements have a significant impact. Recently, efforts have been focused on maximizing the performance of classification neural networks; nevertheless, their performance is limited by the quality of the input data. Monochrome linear backlight systems are most commonly used for online inspections of food packaging thermo-sealing zones. These simple acquisition systems fit the high cadence of the production lines imposed by the market demand. Nevertheless, they provide a limited amount of data, which negatively impacts classification algorithm training. A desired situation would be one where data quality is maximized in terms of obtaining the key information to detect defects while maintaining a fast working pace. This work presents a backlight hyperspectral imaging system designed and implemented replicating an industrial environment to better understand the relationship between visual data quality and spectral illumination range for a variety of packed food products. Furthermore, results led to the identification of advantageous spectral bands that significantly enhance image quality, providing clearer detection of defects.

Keywords: artificial vision, food packaging, hyperspectral imaging, image acquisition, quality control

Procedia PDF Downloads 23

Authors: Mehrnoosh Abouzari, Shahrokh Sahraei

Abstract:

Today, with the increase in quantity and quality and variety of crimes, the discussion of crime prevention has faced a serious challenge that human resources alone and with traditional methods will not be effective. One of the developments in the modern world is the presence of artificial intelligence in various fields, including criminal law. In fact, the use of artificial intelligence in criminal investigations and fighting crime is a necessity in today's world. The use of artificial intelligence is far beyond and even separate from other technologies in the struggle against crime. Second, its application in criminal science is different from the discussion of prevention and it comes to the prediction of crime. Crime prevention in terms of the three factors of the offender, the offender and the victim, following a change in the conditions of the three factors, based on the perception of the criminal being wise, and therefore increasing the cost and risk of crime for him in order to desist from delinquency or to make the victim aware of self-care and possibility of exposing him to danger or making it difficult to commit crimes. While the presence of artificial intelligence in the field of combating crime and social damage and dangers, like an all-seeing eye, regardless of time and place, it sees the future and predicts the occurrence of a possible crime, thus prevent the occurrence of crimes. The purpose of this article is to collect and analyze the studies conducted on the use of artificial intelligence in predicting and preventing crime. How capable is this technology in predicting crime and preventing it? The results have shown that the artificial intelligence technologies in use are capable of predicting and preventing crime and can find patterns in the data set. find large ones in a much more efficient way than humans. In crime prediction and prevention, the term artificial intelligence can be used to refer to the increasing use of technologies that apply algorithms to large sets of data to assist or replace police. The use of artificial intelligence in our debate is in predicting and preventing crime, including predicting the time and place of future criminal activities, effective identification of patterns and accurate prediction of future behavior through data mining, machine learning and deep learning, and data analysis, and also the use of neural networks. Because the knowledge of criminologists can provide insight into risk factors for criminal behavior, among other issues, computer scientists can match this knowledge with the datasets that artificial intelligence uses to inform them.

Keywords: artificial intelligence, criminology, crime, prevention, prediction

Procedia PDF Downloads 77

Authors: Shalaka Kadam, Dinesh Apte, Sagar Mainkar

Abstract:

The life cycle of products is becoming shorter and shorter due to increased competition in market, shorter product development time and increased product diversity. Short life cycles are normal in retail industry, style business, entertainment media, and telecom and semiconductor industry. The subject of accurate forecasting for demand of short lifecycle products is of special enthusiasm for many researchers and organizations. Due to short life cycle of products the amount of historical data that is available for forecasting is very minimal or even absent when new or modified products are launched in market. The companies dealing with such products want to increase the accuracy in demand forecasting so that they can utilize the full potential of the market at the same time do not oversupply. This provides the challenge to develop a forecasting model that can forecast accurately while handling large variations in data and consider the complex relationships between various parameters of data. Many statistical models have been proposed in literature for forecasting time series data. Traditional time series forecasting models do not work well for short life cycles due to lack of historical data. Also artificial neural networks (ANN) models are very time consuming to perform forecasting. We have studied the existing models that are used for forecasting and their limitations. This work proposes an effective and powerful forecasting approach for short life cycle time series forecasting. We have proposed an approach which takes into consideration different scenarios related to data availability for short lifecycle products. We then suggest a methodology which combines statistical analysis with structured judgement. Also the defined approach can be applied across domains. We then describe the method of creating a profile from analogous products. This profile can then be used for forecasting products with historical data of analogous products. We have designed an application which combines data, analytics and domain knowledge using point-and-click technology. The forecasting results generated are compared using MAPE, MSE and RMSE error scores. Conclusion: Based on the results it is observed that no one approach is sufficient for short life-cycle forecasting and we need to combine two or more approaches for achieving the desired accuracy.

Keywords: forecast, short life cycle product, structured judgement, time series

Procedia PDF Downloads 360

Authors: Mohammed Al Sulaimani, Hamad Al Manhi

Abstract:

With the development of Remote Sensing technology, the resolution of optical Remote Sensing images has greatly improved, and images have become largely available. Numerous detectors have been developed for detecting different types of objects. In the past few years, Remote Sensing has benefited a lot from deep learning, particularly Deep Convolution Neural Networks (CNNs). Deep learning holds great promise to fulfill the challenging needs of Remote Sensing and solving various problems within different fields and applications. The use of Unmanned Aerial Systems in acquiring Aerial Photos has become highly used and preferred by most organizations to support their activities because of their high resolution and accuracy, which make the identification and detection of very small features much easier than Satellite Images. And this has opened an extreme era of Deep Learning in different applications not only in feature extraction and prediction but also in analysis. This work addresses the capacity of Machine Learning and Deep Learning in detecting and extracting Oil Leaks from Flowlines (Onshore) using High-Resolution Aerial Photos which have been acquired by UAS fixed with RGB Sensor to support early detection of these leaks and prevent the company from the leak’s losses and the most important thing environmental damage. Here, there are two different approaches and different methods of DL have been demonstrated. The first approach focuses on detecting the Oil Leaks from the RAW Aerial Photos (not processed) using a Deep Learning called Single Shoot Detector (SSD). The model draws bounding boxes around the leaks, and the results were extremely good. The second approach focuses on detecting the Oil Leaks from the Ortho-mosaiced Images (Georeferenced Images) by developing three Deep Learning Models using (MaskRCNN, U-Net and PSP-Net Classifier). Then, post-processing is performed to combine the results of these three Deep Learning Models to achieve a better detection result and improved accuracy. Although there is a relatively small amount of datasets available for training purposes, the Trained DL Models have shown good results in extracting the extent of the Oil Leaks and obtaining excellent and accurate detection.

Keywords: GIS, remote sensing, oil leak detection, machine learning, aerial photos, unmanned aerial systems

Procedia PDF Downloads 34

Authors: M. Cabrera, W. Tizani, J. Ninic, F. Wang

Abstract:

Open section beam to concrete-filled tubular column structures has been increasingly utilized in construction over the past few decades due to their enhanced structural performance, as well as economic and architectural advantages. However, the use of this configuration in construction is limited due to the difficulties in connecting the structural members as there is no access to the inner part of the tube to install standard bolts. Blind-bolted systems are a relatively new approach to overcome this limitation as they only require access to one side of the tubular section to tighten the bolt. The performance of these connections in concrete-filled steel tubular sections remains uncharacterized due to the complex interactions between concrete, bolt, and steel section. Over the last years, research in structural performance has moved to a more sophisticated and efficient approach consisting of machine learning algorithms to generate metamodels. This method reduces the need for developing complex, and computationally expensive finite element models, optimizing the search for desirable design variables. Metamodels generated by a data fusion approach use numerical and experimental results by combining multiple models to capture the dependency between the simulation design variables and connection performance, learning the relations between different design parameters and predicting a given output. Fully characterizing this connection will transform high-rise and multistorey construction by means of the introduction of design guidance for moment-resisting blind-bolted connections, which is currently unavailable. This paper presents a review of the steps taken to develop metamodels generated by means of artificial neural network algorithms which predict the connection stress and stiffness based on the design parameters when using Extended Hollo-Bolt blind bolts. It also provides consideration of the failure modes and mechanisms that contribute to the deformability as well as the feasibility of achieving blind-bolted rigid connections when using the blind fastener.

Keywords: blind-bolted connections, concrete-filled tubular structures, finite element analysis, metamodeling

Procedia PDF Downloads 158

Authors: Ahmedou Moulaye Idriss, Tfeil Yahya, Tamas Ungi, Gabor Fichtinger

Abstract:

Image-guided therapy (IGT) plays a crucial role in minimally invasive procedures for liver interventions. Accurate segmentation of the liver and precise needle placement is essential for successful interventions such as needle aspiration. In this study, we propose a deep learning-based liver 3D slicer designed to enhance segmentation accuracy and facilitate needle aspiration procedures. The developed 3D slicer leverages state-of-the-art convolutional neural networks (CNNs) for automatic liver segmentation in medical images. The CNN model is trained on a diverse dataset of liver images obtained from various imaging modalities, including computed tomography (CT) and magnetic resonance imaging (MRI). The trained model demonstrates robust performance in accurately delineating liver boundaries, even in cases with anatomical variations and pathological conditions. Furthermore, the 3D slicer integrates advanced image registration techniques to ensure accurate alignment of preoperative images with real-time interventional imaging. This alignment enhances the precision of needle placement during aspiration procedures, minimizing the risk of complications and improving overall intervention outcomes. To validate the efficacy of the proposed deep learning-based 3D slicer, a comprehensive evaluation is conducted using a dataset of clinical cases. Quantitative metrics, including the Dice similarity coefficient and Hausdorff distance, are employed to assess the accuracy of liver segmentation. Additionally, the performance of the 3D slicer in guiding needle aspiration procedures is evaluated through simulated and clinical interventions. Preliminary results demonstrate the effectiveness of the developed 3D slicer in achieving accurate liver segmentation and guiding needle aspiration procedures with high precision. The integration of deep learning techniques into the IGT workflow shows great promise for enhancing the efficiency and safety of liver interventions, ultimately contributing to improved patient outcomes.

Keywords: deep learning, liver segmentation, 3D slicer, image guided therapy, needle aspiration

Procedia PDF Downloads 52

Authors: Teris Cheung, Joyce Yuen Ting Lam, Kwan Hin Fong, Yuen Shan Ho, Tim Man Ho Li, Andy Choi-Yeung Tse, Cheng-Ta Li, Calvin Pak-Wing Cheng, Roland Beisteiner

Abstract:

Transcranial pulse stimulation (TPS) is a non-intrusive brain stimulation technology that has been proven effective in older adults with mild neurocognitive disorders and adults with major depressive disorder. Given these robust evidences, TPS might be an adjunct treatment options in neuropsychiatric disorders, for example, autism spectrum disorder (ASD) – which is a common neurodevelopmental disorder in children. This trial aimed to investigate the effects of TPS on right temporoparietal junction, a key node for social cognition for Autism Spectrum Disorder (ASD), and to examine the association between TPS, executive functions and social functions. Design: This trial adopted a two-armed (verum TPS group vs. sham TPS group), double-blinded, randomized, sham-controlled design. Sampling: 32 subjects aged between 12 and 17, diagnosed with ASD were recruited. All subjects were computerized randomized into either verum TPS group or the sham TPS group on a 1:1 ratio. All subjects undertook functional MRI before and after the TPS interventions. Intervention: Six 30-min TPS sessions were administered to subjects in 2 weeks’ time on alternate days assessing neural connectivity changes. Baseline measurements and post-TPS evaluation of the ASD symptoms, executive functions, and social functions were conducted. Participants were followed up at 2-weeks, at 1-month and 3-month, assessing the short-and long-term sustainability of the TPS intervention. Data analysis: Generalized Estimating Equations with repeated measures were used to analyze the group and time difference. Missing data were managed by multiple imputations. The level of significance was set at p < 0.05. To our best knowledge, this is the first study evaluating the efficacy and safety of TPS among adolescents with ASD in Hong Kong and nationwide. Results emerging from this study will develop insight on whether TPS can be used as an adjunct treatment on ASD in neuroscience and clinical psychiatry. Clinical Trial Registration: ClinicalTrials.gov, identifier: NCT05408793.

Keywords: adolescents, autism spectrum disorder, neuromodulation, rct, transcranial pulse stimulation

Procedia PDF Downloads 74

Authors: Djeddi Khaled, Houssou Hind, Miloudi Abdellatif, Rabah Siham

Abstract:

In the field of veterinary medicine, there is a growing application of artificial intelligence (AI) for diagnosing bovine mastitis, a prevalent inflammatory disease in dairy cattle. AI technologies, such as automated milking systems, have streamlined the assessment of key metrics crucial for managing cow health during milking and identifying prevalent diseases, including mastitis. These automated milking systems empower farmers to implement automatic mastitis detection by analyzing indicators like milk yield, electrical conductivity, fat, protein, lactose, blood content in the milk, and milk flow rate. Furthermore, reports highlight the integration of somatic cell count (SCC), thermal infrared thermography, and diverse systems utilizing statistical models and machine learning techniques, including artificial neural networks, to enhance the overall efficiency and accuracy of mastitis detection. According to a review of 15 publications, machine learning technology can predict the risk and detect mastitis in cattle with an accuracy ranging from 87.62% to 98.10% and sensitivity and specificity ranging from 84.62% to 99.4% and 81.25% to 98.8%, respectively. Additionally, machine learning algorithms and microarray meta-analysis are utilized to identify mastitis genes in dairy cattle, providing insights into the underlying functional modules of mastitis disease. Moreover, AI applications can assist in developing predictive models that anticipate the likelihood of mastitis outbreaks based on factors such as environmental conditions, herd management practices, and animal health history. This proactive approach supports farmers in implementing preventive measures and optimizing herd health. By harnessing the power of artificial intelligence, the diagnosis of bovine mastitis can be significantly improved, enabling more effective management strategies and ultimately enhancing the health and productivity of dairy cattle. The integration of artificial intelligence presents valuable opportunities for the precise and early detection of mastitis, providing substantial benefits to the dairy industry.

Keywords: artificial insemination, automatic milking system, cattle, machine learning, mastitis

Procedia PDF Downloads 66

Authors: Shankha Sanyal, Archi Banerjee, Sayan Biswas, Sourya Sengupta, Sayan Nag, Ranjan Sengupta, Dipak Ghosh

Abstract:

The term ‘gestalt’ has been widely used in the field of psychology which defined the perception of human mind to group any object not in part but as a 'unified' whole. Music, in general, is polyphonic - i.e. a combination of a number of pure tones (frequencies) mixed together in a manner that sounds harmonious. The study of human brain response due to different frequency groups of the acoustic signal can give us an excellent insight regarding the neural and functional architecture of brain functions. Hence, the study of music cognition using neuro-biosensors is becoming a rapidly emerging field of research. In this work, we have tried to analyze the effect of different frequency bands of music on the various frequency rhythms of human brain obtained from EEG data. Four widely popular Rabindrasangeet clips were subjected to Wavelet Transform method for extracting five resonant frequency bands from the original music signal. These frequency bands were initially analyzed with Detrended/Adaptive Fractal analysis (DFA/AFA) methods. A listening test was conducted on a pool of 100 respondents to assess the frequency band in which the music becomes non-recognizable. Next, these resonant frequency bands were presented to 20 subjects as auditory stimulus and EEG signals recorded simultaneously in 19 different locations of the brain. The recorded EEG signals were noise cleaned and subjected again to DFA/AFA technique on the alpha, theta and gamma frequency range. Thus, we obtained the scaling exponents from the two methods in alpha, theta and gamma EEG rhythms corresponding to different frequency bands of music. From the analysis of music signal, it is seen that loss of recognition is proportional to the loss of long range correlation in the signal. From the EEG signal analysis, we obtain frequency specific arousal based response in different lobes of brain as well as in specific EEG bands corresponding to musical stimuli. In this way, we look to identify a specific frequency band beyond which the music becomes non-recognizable and below which in spite of the absence of other bands the music is perceivable to the audience. This revelation can be of immense importance when it comes to the field of cognitive music therapy and researchers of creativity.

Keywords: AFA, DFA, EEG, gestalt in music, Hurst exponent

Procedia PDF Downloads 332

Authors: Davide Palma, Dimitra Papagiannaki, Marco Minella, Manuel Lai, Rita Binetti, Claire Richard

Abstract:

Modern analytical technologies allow us to detect water contaminants at trace and ultra-trace concentrations highlighting how a large number of organic compounds is not efficiently abated by most wastewater treatment facilities relying on biological processes; we usually refer to these micropollutants as contaminants of emerging concern (CECs). The availability of reliable end effective technologies, able to guarantee the high standards of water quality demanded by legislators worldwide, has therefore become a primary need. In this context, water plasma stands out among developing technologies as it is extremely effective in the abatement of numerous classes of pollutants, cost-effective, and environmentally friendly. In this work, a custom-built non-thermal pulsed plasma discharge generator was used to abate the concentration of selected CECs in the water samples. Samples were treated in a 50 mL pyrex reactor using two different types of plasma discharge occurring at the surface of the treated solution or, underwater, working with positive polarity. The distance between the tips of the electrodes determined where the discharge was formed: underwater when the distance was < 2mm, at the water surface when the distance was > 2 mm. Peak voltage was in the 100-130kV range with typical current values of 20-40 A. The duration of the pulse was 500 ns, and the frequency of discharge could be manually set between 5 and 45 Hz. Treatment of 100 µM diclofenac solution in MilliQ water, with a pulse frequency of 17Hz, revealed that surface discharge was more efficient in the degradation of diclofenac that was no longer detectable after 6 minutes of treatment. Over 30 minutes were required to obtain the same results with underwater discharge. These results are justified by the higher rate of H₂O₂ formation (21.80 µmolL⁻¹min⁻¹ for surface discharge against 1.20 µmolL⁻¹min⁻¹ for underwater discharge), larger discharge volume and UV light emission, high rate of ozone and NOx production (up to 800 and 1400 ppb respectively) observed when working with surface discharge. Then, the surface discharge was used for the treatment of the three selected perfluoroalkyl compounds, namely, perfluorooctanoic acid (PFOA), perfluorohexanoic acid (PFHxA), and pefluorooctanesulfonic acid (PFOS) both individually and in mixture, in ultrapure and groundwater matrices with initial concentration of 1 ppb. In both matrices, PFOS exhibited the best degradation reaching complete removal after 30 min of treatment (degradation rate 0.107 min⁻¹ in ultrapure water and 0.0633 min⁻¹ in groundwater), while the degradation rate of PFOA and PFHxA was slower of around 65% and 80%, respectively. Total nitrogen (TN) measurements revealed levels up to 45 mgL⁻¹h⁻¹ in water samples treated with surface discharge, while, in analogous samples treated with underwater discharge, TN increase was 5 to 10 times lower. These results can be explained by the significant NOx concentrations (over 1400 ppb) measured above functioning reactor operating with superficial discharge; rapid NOx hydrolysis led to nitrates accumulation in the solution explaining the observed evolution of TN values. Ionic chromatography measures confirmed that the vast majority of TN was under the form of nitrates. In conclusion, non-thermal pulsed plasma discharge, obtained with a custom-built generator, was proven to effectively degrade diclofenac in water matrices confirming the potential interest of this technology for wastewater treatment. The surface discharge was proven to be more effective in CECs removal due to the high rate of formation of H₂O₂, ozone, reactive radical species, and strong UV light emission. Furthermore, nitrates enriched water obtained after treatment could be an interesting added-value product to be used as fertilizer in agriculture. Acknowledgment: This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 765860.

Keywords: CECs removal, nitrogen fixation, non-thermal plasma, water treatment

Procedia PDF Downloads 121

Authors: C. J. Rossouw, T. I. van Niekerk

Abstract:

The global manufacturing industry is utilizing the internet and cloud-based services to further explore the anatomy and optimize manufacturing processes in support of the movement into the Fourth Industrial Revolution (4IR). The 4IR from a third world and African perspective is hindered by the fact that many manufacturing systems that were developed in the third industrial revolution are not inherently equipped to utilize the internet and services of the 4IR, hindering the progression of third world manufacturing industries into the 4IR. This research focuses on the development of a non-invasive and cost-effective cyber-physical IoT system that will exploit a machine’s vibration to expose semantic characteristics in the manufacturing process and utilize these results through a real-time cloud-based machine condition monitoring system with the intention to optimize the system. A microcontroller-based IoT sensor was designed to acquire a machine’s mechanical vibration data, process it in real-time, and transmit it to a cloud-based platform via Wi-Fi and the internet. Time-frequency Fourier analysis was applied to the vibration data to form an image representation of the machine’s behaviour. This data was used to train a Convolutional Neural Network (CNN) to learn semantic characteristics in the machine’s behaviour and relate them to a state of operation. The same data was also used to train a Convolutional Autoencoder (CAE) to detect anomalies in the data. Real-time edge-based artificial intelligence was achieved by deploying the CNN and CAE on the sensor to analyse the vibration. A cloud platform was deployed to visualize the vibration data and the results of the CNN and CAE in real-time. The cyber-physical IoT system was deployed on a semi-automated metal granulation machine with a set of trained machine learning models. Using a single sensor, the system was able to accurately visualize three states of the machine’s operation in real-time. The system was also able to detect a variance in the material being granulated. The research demonstrates how non-IoT manufacturing systems can be equipped with edge-based artificial intelligence to establish a remote machine condition monitoring system.

Keywords: IoT, cyber-physical systems, artificial intelligence, manufacturing, vibration analytics, continuous machine condition monitoring

Procedia PDF Downloads 88

Authors: Chamika Chiran Perera, Dananjaya Perera, Chirath Dasanayake, Banuka Athuraliya

Abstract:

Male infertility is a major problem in the world, and it is a neglected and sensitive health issue in Sri Lanka. It can be determined by analyzing human semen samples. Sperm motility is one of many factors that can evaluate male’s fertility potential. In Sri Lanka, this analysis is performed manually. Manual methods are time consuming and depend on the person, but they are reliable and it can depend on the expert. Machine learning and deep learning technologies are currently being investigated to automate the spermatozoa motility analysis, and these methods are unreliable. These automatic methods tend to produce false positive results and false detection. Current automatic methods support different techniques, and some of them are very expensive. Due to the geographical variance in spermatozoa characteristics, current automatic methods are not reliable for motility analysis in Sri Lanka. The suggested system, DeepMotile, is to explore a method to analyze motility of human spermatozoa automatically and present it to the andrology laboratories to overcome current issues. DeepMotile is a novel deep learning method for analyzing spermatozoa motility parameters in the Sri Lankan population. To implement the current approach, Sri Lanka patient data were collected anonymously as a dataset, and glass slides were used as a low-cost technique to analyze semen samples. Current problem was identified as microscopic object detection and tackling the problem. YOLOv5 was customized and used as the object detector, and it achieved 94 % mAP (mean average precision), 86% Precision, and 90% Recall with the gathered dataset. StrongSORT was used as the object tracker, and it was validated with andrology experts due to the unavailability of annotated ground truth data. Furthermore, this research has identified many potential ways for further investigation, and andrology experts can use this system to analyze motility parameters with realistic accuracy.

Keywords: computer vision, deep learning, convolutional neural networks, multi-target tracking, microscopic object detection and tracking, male infertility detection, motility analysis of human spermatozoa

Procedia PDF Downloads 107

Authors: Sara Tokhi Arab, Rozo Noguchi, Tofeal Ahamed

Abstract:

Drought is a natural disaster that occurs in a region due to a lack of precipitation and high temperatures over a continuous period or in a single season as a consequence of climate change. Precipitation deficits and prolonged high temperatures mostly affect the agricultural sector, water resources, socioeconomics, and the environment. Consequently, it causes agricultural product loss, food shortage, famines, migration, and natural resources degradation in a region. Agriculture is the first sector affected by drought. Therefore, it is important to develop an agricultural drought risk and loss assessment to mitigate the drought impact in the agriculture sector. In this context, the main purpose of this study was to assess yield loss using composite drought indices in the drought-affected vineyards. In this study, the CDI was developed for the years 2016 to 2020 by comprising five indices: the vegetation condition index (VCI), temperature condition index (TCI), deviation of NDVI from the long-term mean (NDVI DEV), normalized difference moisture index (NDMI) and precipitation condition index (PCI). Moreover, the quantitative principal component analysis (PCA) approach was used to assign a weight for each input parameter, and then the weights of all the indices were combined into one composite drought index. Finally, Bayesian regularized artificial neural networks (BRANNs) were used to evaluate the yield variation in each affected vineyard. The composite drought index result indicated the moderate to severe droughts were observed across the Kabul Province during 2016 and 2018. Moreover, the results showed that there was no vineyard in extreme drought conditions. Therefore, we only considered the severe and moderated condition. According to the BRANNs results R=0.87 and R=0.94 in severe drought conditions for the years of 2016 and 2018 and the R= 0.85 and R=0.91 in moderate drought conditions for the years of 2016 and 2018, respectively. In the Kabul Province within the two years drought periods, there was a significate deficit in the vineyards. According to the findings, 2018 had the highest rate of loss almost -7 ton/ha. However, in 2016 the loss rates were about – 1.2 ton/ha. This research will support stakeholders to identify drought affect vineyards and support farmers during severe drought.

Keywords: grapes, composite drought index, yield loss, satellite remote sensing

Procedia PDF Downloads 158

Authors: Mohammad Khazaei, Reza Rostami, Hasan Gharayagh Zandi, Rouhollah Basatnia, Mahbubeh Ghayour Najafabadi

Abstract:

Objective: Methylphenidate has been demonstrated to enhance attention and cognitive processes, and placebo treatments have also been found to improve attention and cognitive processes. Additionally, methylphenidate may have positive effects on motion perception and sports performance. Nevertheless, additional research is needed to fully comprehend the neural mechanisms underlying the effects of methylphenidate and placebo on cognitive and motor functions. Methods: In this randomized controlled trial, 18 young semi-professional basketball players aged 18-23 years were randomly and equally assigned to either a Ritalin or Placebo group. The participants performed 20 consecutive free throws; their scores were recorded on a 0-3 scale. The participants’ brain activity was recorded using electroencephalography (EEG) for 5 minutes seated with their eyes closed. The Ritalin group received a 10 mg dose of methylphenidate, while the Placebo group received a 10mg dose of placebo. The EEG was obtained 90 minutes after the drug was administere Results: There was no significant difference in the absolute power of brain waves between the pre-test and post-tests in the Placebo group. However, in the Ritalin group, a significant difference in the absolute power of brain waves was observed in the Theta band (5-6 Hz) and Beta band (21-30 Hz) between pre- and post-tests in Fp2, F8, and Fp1. In these areas, the absolute power of Beta waves was higher during the post-test than during the pre-test. The Placebo group showed a more significant difference in free throw scores than the Ritalin group. Conclusions: In conclusion, these results suggest that Ritalin effect on brain activity in areas associated with attention and cognitive processes, as well as improve basketball free throws. However, there was no significant placebo effect on brain activity performance, but it significantly affected the improvement of free throws. Further research is needed to fully understand the effects of methylphenidate and placebo on cognitive and motor functions.

Keywords: methylphenidate, placebo effect, electroencephalography, basketball free throw

Procedia PDF Downloads 80

Authors: Sarayu Vanga, Jorge Galeano-Cabral, Kaya Wei

Abstract:

Cerebral hypoxia refers to a condition in which there is a decrease in oxygen supply to the brain. Patients suffering from this condition experience a decrease in their body temperature. While there isn't any cure to treat cerebral hypoxia as of date, certain procedures are utilized to help aid in the treatment of the condition. Regulating the body temperature is an example of one of those procedures. Hypoxia is well known to reduce the body temperature of mammals, although the neural origins of this response remain uncertain. In order to speed recovery from this condition, it is necessary to maintain a stable body temperature. In this study, we present an approach to regulating body temperature for patients who suffer from cerebral hypoxia or other similar conditions. After a thorough literature study, we propose the use of thermoelectric blankets, which are temperature-controlled thermal blankets based on thermoelectric devices. These blankets are capable of heating up and cooling down the patient to stabilize body temperature. This feature is possible through the reversible effect that thermoelectric devices offer while behaving as a thermal sensor, and it is an effective way to stabilize temperature. Thermoelectricity is the direct conversion of thermal to electrical energy and vice versa. This effect is now known as the Seebeck effect, and it is characterized by the Seebeck coefficient. In such a configuration, the device has cooling and heating sides with temperatures that can be interchanged by simply switching the direction of the current input in the system. This design integrates various aspects, including a humidifier, ventilation machine, IV-administered medication, air conditioning, circulation device, and a body temperature regulation system. The proposed design includes thermocouples that will trigger the blanket to increase or decrease a set temperature through a medical temperature sensor. Additionally, the proposed design allows an efficient way to control fluctuations in body temperature while being cost-friendly, with an expected cost of 150 dollars. We are currently working on developing a prototype of the design to collect thermal and electrical data under different conditions and also intend to perform an optimization analysis to improve the design even further. While this proposal was developed for treating cerebral hypoxia, it can also aid in the treatment of other related conditions, as fluctuations in body temperature appear to be a common symptom that patients have for many illnesses.

Keywords: body temperature regulation, cerebral hypoxia, thermoelectric, blanket design

Procedia PDF Downloads 161

Authors: Shan Zhang, Peter Suechting

Abstract:

Predicting trade patterns is critical to decision-making in public and private domains, especially in the current context of trade disputes among major economies. In the past, U.S. recycling has relied heavily on strong demand for recyclable materials overseas. However, starting in 2017, a series of new recycling policies (bans and higher inspection standards) was enacted by multiple countries that were the primary importers of recyclables from the U.S. prior to that point. As the global trade flow of recycling shifts, some new importers, mostly developing countries in South and Southeast Asia, have been overwhelmed by the sheer quantities of scrap materials they have received. As the leading exporter of recyclable materials, the U.S. now has a pressing need to build its recycling industry domestically. With respect to the global trade in scrap materials used for recycling, the interest in this paper is (1) predicting how the export of recyclable materials from the U.S. might vary over time, and (2) predicting how international trade flows for recyclables might change in the future. Focusing on three major recyclable materials with a history of trade, this study uses data-driven and machine learning (ML) algorithms---supervised (shrinkage and tree methods) and unsupervised (neural network method)---to decipher the international trade pattern of recycling. Forecasting the potential trade values of recyclables in the future could help importing countries, to which those materials will shift next, to prepare related trade policies. Such policies can assist policymakers in minimizing negative environmental externalities and in finding the optimal amount of recyclables needed by each country. Such forecasts can also help exporting countries, like the U.S understand the importance of healthy domestic recycling industry. The preliminary result suggests that gravity models---in addition to particular selection macroeconomic predictor variables--are appropriate predictors of the total export value of recyclables. With the inclusion of variables measuring aspects of the political conditions (trade tariffs and bans), predictions show that recyclable materials are shifting from more policy-restricted countries to less policy-restricted countries in international recycling trade. Those countries also tend to have high manufacturing activities as a percentage of their GDP.

Keywords: environmental economics, machine learning, recycling, international trade

Procedia PDF Downloads 170

Authors: Yanjun Chen, Chuanping Hu, Jie Shao, Lin Mei, Chongyang Zhang

Abstract:

Traditional image classification is based on encoding scheme (e.g. Fisher Vector, Vector of Locally Aggregated Descriptor) with low-level image features (e.g. SIFT, HoG). Compared to these low-level local features, deep convolutional features obtained at the mid-level layer of convolutional neural networks (CNN) have richer information but lack of geometric invariance. For scene classification, there are scattered objects with different size, category, layout, number and so on. It is crucial to find the distinctive objects in scene as well as their co-occurrence relationship. In this paper, we propose a method to take advantage of both deep convolutional features and the traditional encoding scheme while taking object-centric and scene-centric information into consideration. First, to exploit the object-centric and scene-centric information, two CNNs that trained on ImageNet and Places dataset separately are used as the pre-trained models to extract deep convolutional features at multiple scales. This produces dense local activations. By analyzing the performance of different CNNs at multiple scales, it is found that each CNN works better in different scale ranges. A scale-wise CNN adaption is reasonable since objects in scene are at its own specific scale. Second, a fisher kernel is applied to aggregate a global representation at each scale and then to merge into a single vector by using a post-processing method called scale-wise normalization. The essence of Fisher Vector lies on the accumulation of the first and second order differences. Hence, the scale-wise normalization followed by average pooling would balance the influence of each scale since different amount of features are extracted. Third, the Fisher vector representation based on the deep convolutional features is followed by a linear Supported Vector Machine, which is a simple yet efficient way to classify the scene categories. Experimental results show that the scale-specific feature extraction and normalization with CNNs trained on object-centric and scene-centric datasets can boost the results from 74.03% up to 79.43% on MIT Indoor67 when only two scales are used (compared to results at single scale). The result is comparable to state-of-art performance which proves that the representation can be applied to other visual recognition tasks.

Keywords: deep convolutional features, Fisher Vector, multiple scales, scale-specific normalization

Procedia PDF Downloads 333

Authors: Zachary Huffman, Joana Rocha

Abstract:

Wall-pressure fluctuations induced by the turbulent boundary layer (TBL) developed over aircraft are a significant source of aircraft cabin noise. Since the power spectral density (PSD) of these pressure fluctuations is directly correlated with the amount of sound radiated into the cabin, the development of accurate empirical models that predict the PSD has been an important ongoing research topic. The sound emitted can be represented from the pressure fluctuations term in the Reynoldsaveraged Navier-Stokes equations (RANS). Therefore, early TBL empirical models (including those from Lowson, Robertson, Chase, and Howe) were primarily derived by simplifying and solving the RANS for pressure fluctuation and adding appropriate scales. Most subsequent models (including Goody, Efimtsov, Laganelli, Smol’yakov, and Rackl and Weston models) were derived by making modifications to these early models or by physical principles. Overall, these models have had varying levels of accuracy, but, in general, they are most accurate under the specific Reynolds and Mach numbers they were developed for, while being less accurate under other flow conditions. Despite this, recent research into the possibility of using alternative methods for deriving the models has been rather limited. More recent studies have demonstrated that an artificial neural network model was more accurate than traditional models and could be applied more generally, but the accuracy of other machine learning techniques has not been explored. In the current study, an original model is derived using a stepwise regression algorithm in the statistical programming language R, and TBL wall-pressure fluctuations PSD data gathered at the Carleton University wind tunnel. The theoretical advantage of a stepwise regression approach is that it will automatically filter out redundant or uncorrelated input variables (through the process of feature selection), and it is computationally faster than machine learning. The main disadvantage is the potential risk of overfitting. The accuracy of the developed model is assessed by comparing it to independently sourced datasets.

Keywords: aircraft noise, machine learning, power spectral density models, regression models, turbulent boundary layer wall-pressure fluctuations

Procedia PDF Downloads 135

Authors: Badriah Alayidi, Emad Alyahya

Abstract:

Background: Most people with Multiple Sclerosis (MS) report worsening balance as the condition progresses. Poor balance control is also well known to be a significant risk factor for both falling and fear of falling. The increased risk of falls with disease progression thus makes balance control an essential target of gait rehabilitation amongst people with MS. Intervention programs have developed various methods to improve balance control, and accumulating evidence suggests that exercise programs may help people with MS improve their balance. Among these methods, virtual reality (VR) is growing in popularity as a balance-training technique owing to its potential benefits, including better compliance and greater user happiness. However, it is not clear if a VR environment will induce different balance control mechanisms in MS as compared to healthy individuals or traditional environments. Therefore, this study aims to examine how individuals with MS control their balance in a VR setting. Methodology: The proposed study takes an empirical approach to estimate and determine the role of balance response in persons with MS using a VR environment. It will use primary data collected through patient observations, physiological and biomechanical evaluation of balance, and data analysis. Results: The preliminary systematic review and meta-analysis indicated that there was variability in terms of the outcome assessing balance response in people with MS. The preliminary results of these assessments have the potential to provide essential indicators of the progression of MS and contribute to the individualization of treatment and evaluation of the interventions’ effectiveness. The literature describes patients who have had the opportunity to experiment in VR settings and then used what they have learned in the real world, suggesting that this VR setting could be more appealing than conditional settings. The findings of the proposed study will be beneficial in estimating and determining the effect of VR on balance control in persons with MS. In previous studies, VR was shown to be an interesting approach to neurological rehabilitation, but more data are needed to support this approach in MS. Conclusions: The proposed study enables an assessment of balance and evaluations of a variety of physiological implications related to neural activity as well as biomechanical implications related to movement analysis.

Keywords: multiple sclerosis, virtual reality, postural control, balance

Procedia PDF Downloads 76

Authors: Man-Yun Liu, Emily Chia-Yu Su

Abstract:

Alzheimer's disease (AD) is the public health crisis of the 21st century. AD is a degenerative brain disease and the most common cause of dementia, a costly disease on the healthcare system. Unfortunately, the cause of AD is poorly understood, furthermore; the treatments of AD so far can only alleviate symptoms rather cure or stop the progress of the disease. Currently, there are several ways to diagnose AD; medical imaging can be used to distinguish between AD, other dementias, and early onset AD, and cerebrospinal fluid (CSF). Compared with other diagnostic tools, blood (plasma) test has advantages as an approach to population-based disease screening because it is simpler, less invasive also cost effective. In our study, we used blood biomarkers dataset of The Alzheimer’s disease Neuroimaging Initiative (ADNI) which was funded by National Institutes of Health (NIH) to do data analysis and develop a prediction model. We used independent analysis of datasets to identify plasma protein biomarkers predicting early onset AD. Firstly, to compare the basic demographic statistics between the cohorts, we used SAS Enterprise Guide to do data preprocessing and statistical analysis. Secondly, we used logistic regression, neural network, decision tree to validate biomarkers by SAS Enterprise Miner. This study generated data from ADNI, contained 146 blood biomarkers from 566 participants. Participants include cognitive normal (healthy), mild cognitive impairment (MCI), and patient suffered Alzheimer’s disease (AD). Participants’ samples were separated into two groups, healthy and MCI, healthy and AD, respectively. We used the two groups to compare important biomarkers of AD and MCI. In preprocessing, we used a t-test to filter 41/47 features between the two groups (healthy and AD, healthy and MCI) before using machine learning algorithms. Then we have built model with 4 machine learning methods, the best AUC of two groups separately are 0.991/0.709. We want to stress the importance that the simple, less invasive, common blood (plasma) test may also early diagnose AD. As our opinion, the result will provide evidence that blood-based biomarkers might be an alternative diagnostics tool before further examination with CSF and medical imaging. A comprehensive study on the differences in blood-based biomarkers between AD patients and healthy subjects is warranted. Early detection of AD progression will allow physicians the opportunity for early intervention and treatment.

Keywords: Alzheimer's disease, blood-based biomarkers, diagnostics, early detection, machine learning

Procedia PDF Downloads 324

Authors: Brennan Lodge

Abstract:

Amidst a complex regulatory landscape, Retrieval Augmented Generation (RAG) emerges as a transformative tool for Governance Risk and Compliance (GRC) officers. This paper details the application of RAG in synthesizing Large Language Models (LLMs) with external knowledge bases, offering GRC professionals an advanced means to adapt to rapid changes in compliance requirements. While the development for standalone LLM’s (Large Language Models) is exciting, such models do have their downsides. LLM’s cannot easily expand or revise their memory, and they can’t straightforwardly provide insight into their predictions, and may produce “hallucinations.” Leveraging a pre-trained seq2seq transformer and a dense vector index of domain-specific data, this approach integrates real-time data retrieval into the generative process, enabling gap analysis and the dynamic generation of compliance and risk management content. We delve into the mechanics of RAG, focusing on its dual structure that pairs parametric knowledge contained within the transformer model with non-parametric data extracted from an updatable corpus. This hybrid model enhances decision-making through context-rich insights, drawing from the most current and relevant information, thereby enabling GRC officers to maintain a proactive compliance stance. Our methodology aligns with the latest advances in neural network fine-tuning, providing a granular, token-level application of retrieved information to inform and generate compliance narratives. By employing RAG, we exhibit a scalable solution that can adapt to novel regulatory challenges and cybersecurity threats, offering GRC officers a robust, predictive tool that augments their expertise. The granular application of RAG’s dual structure not only improves compliance and risk management protocols but also informs the development of compliance narratives with pinpoint accuracy. It underscores AI’s emerging role in strategic risk mitigation and proactive policy formation, positioning GRC officers to anticipate and navigate the complexities of regulatory evolution confidently.

Keywords: cybersecurity, gen AI, retrieval augmented generation, cybersecurity defense strategies

Procedia PDF Downloads 97

Authors: R. Nandhini, Gaurab Mudbhari

Abstract:

Ranging from the field of health care to self-driving cars, machine learning and deep learning algorithms have revolutionized the field with the proper utilization of images and visual-oriented data. Segmentation, regression, classification, clustering, dimensionality reduction, etc., are some of the Machine Learning tasks that helped Machine Learning and Deep Learning models to become state-of-the-art models for the field where images are key datasets. Among these tasks, classification and clustering are essential but difficult because of the intricate and high-dimensional characteristics of image data. This finding examines and assesses advanced techniques in supervised classification and unsupervised clustering for image datasets, emphasizing the relative efficiency of Convolutional Neural Networks (CNNs), Vision Transformers (ViTs), Deep Embedded Clustering (DEC), and self-supervised learning approaches. Due to the distinctive structural attributes present in images, conventional methods often fail to effectively capture spatial patterns, resulting in the development of models that utilize more advanced architectures and attention mechanisms. In image classification, we investigated both CNNs and ViTs. One of the most promising models, which is very much known for its ability to detect spatial hierarchies, is CNN, and it serves as a core model in our study. On the other hand, ViT is another model that also serves as a core model, reflecting a modern classification method that uses a self-attention mechanism which makes them more robust as this self-attention mechanism allows them to lean global dependencies in images without relying on convolutional layers. This paper evaluates the performance of these two architectures based on accuracy, precision, recall, and F1-score across different image datasets, analyzing their appropriateness for various categories of images. In the domain of clustering, we assess DEC, Variational Autoencoders (VAEs), and conventional clustering techniques like k-means, which are used on embeddings derived from CNN models. DEC, a prominent model in the field of clustering, has gained the attention of many ML engineers because of its ability to combine feature learning and clustering into a single framework and its main goal is to improve clustering quality through better feature representation. VAEs, on the other hand, are pretty well known for using latent embeddings for grouping similar images without requiring for prior label by utilizing the probabilistic clustering method.

Keywords: machine learning, deep learning, image classification, image clustering

Procedia PDF Downloads 17

Authors: Matthew Baucum

Abstract:

With the increasing popularity of fMRI as an experimental method, psychology and neuroscience can greatly benefit from advanced techniques for summarizing and synthesizing large amounts of data from brain imaging studies. One promising avenue is automated meta-analyses, in which natural language processing methods are used to identify the brain regions consistently associated with certain semantic concepts (e.g. “social”, “reward’) across large corpora of studies. This study builds on this approach by demonstrating how, in fMRI meta-analyses, individual voxels can be treated as vectors in a semantic space and evaluated for their “proximity” to terms of interest. In this technique, a low-dimensional semantic space is built from brain imaging study texts, allowing words in each text to be represented as vectors (where words that frequently appear together are near each other in the semantic space). Consequently, each voxel in a brain mask can be represented as a normalized vector sum of all of the words in the studies that showed activation in that voxel. The entire brain mask can then be visualized in terms of each voxel’s proximity to a given term of interest (e.g., “vision”, “decision making”) or collection of terms (e.g., “theory of mind”, “social”, “agent”), as measured by the cosine similarity between the voxel’s vector and the term vector (or the average of multiple term vectors). Analysis can also proceed in the opposite direction, allowing word cloud visualizations of the nearest semantic neighbors for a given brain region. This approach allows for continuous, fine-grained metrics of voxel-term associations, and relies on state-of-the-art “open vocabulary” methods that go beyond mere word-counts. An analysis of over 11,000 neuroimaging studies from an existing meta-analytic fMRI database demonstrates that this technique can be used to recover known neural bases for multiple psychological functions, suggesting this method’s utility for efficient, high-level meta-analyses of localized brain function. While automated text analytic methods are no replacement for deliberate, manual meta-analyses, they seem to show promise for the efficient aggregation of large bodies of scientific knowledge, at least on a relatively general level.

Keywords: FMRI, machine learning, meta-analysis, text analysis

Procedia PDF Downloads 450

Authors: Bahar Hazal Yalçınkaya, Bayram Yılmaz, Mustafa Özilgen

Abstract:

Brain information transmission in the neuronal network occurs in the form of electrical signals. Neural work transmits information between the neurons or neurons and target cells by moving charged particles in a voltage field; a fraction of the energy utilized in this process is dissipated via entropy generation. Exergy loss and entropy generation models demonstrate the inefficiencies of the communication along the dendritic trees. In this study, neurons of 4 different animals were analyzed with one dimensional cable model with N=6 identical dendritic trees and M=3 order of symmetrical branching. Each branch symmetrically bifurcates in accordance with the 3/2 power law in an infinitely long cylinder with the usual core conductor assumptions, where membrane potential is conserved in the core conductor at all branching points. In the model, exergy loss and entropy generation rates are calculated for each branch of equivalent cylinders of electrotonic length (L) ranging from 0.1 to 1.5 for four different dendritic branches, input branch (BI), and sister branch (BS) and two cousin branches (BC-1 & BC-2). Thermodynamic analysis with the data coming from two different cat motoneuron studies show that in both experiments nearly the same amount of exergy is lost while generating nearly the same amount of entropy. Guinea pig vagal motoneuron loses twofold more exergy compared to the cat models and the squid exergy loss and entropy generation were nearly tenfold compared to the guinea pig vagal motoneuron model. Thermodynamic analysis show that the dissipated energy in the dendritic tress is directly proportional with the electrotonic length, exergy loss and entropy generation. Entropy generation and exergy loss show variability not only between the vertebrate and invertebrates but also within the same class. Concurrently, single action potential Na⁺ ion load, metabolic energy utilization and its thermodynamic aspect contributed for squid giant axon and mammalian motoneuron model. Energy demand is supplied to the neurons in the form of Adenosine triphosphate (ATP). Exergy destruction and entropy generation upon ATP hydrolysis are calculated. ATP utilization, exergy destruction and entropy generation showed differences in each model depending on the variations in the ion transport along the channels.

Keywords: ATP utilization, entropy generation, exergy loss, neuronal information transmittance

Procedia PDF Downloads 395

Authors: Vanessa Kummer, Maik Meusel

Abstract:

Real estate demand estimates are typically derived from transaction data. However, in regions with excess demand, transactions are driven by supply and therefore do not indicate what people are actually looking for. To estimate the demand for housing in Switzerland, search subscriptions from all important Swiss real estate platforms are used. These data do, however, suffer from missing information—for example, many users do not specify how many rooms they would like or what price they would be willing to pay. In economic analyses, it is often the case that only complete data is used. Usually, however, the proportion of complete data is rather small which leads to most information being neglected. Also, the data might have a strong distortion if it is complete. In addition, the reason that data is missing might itself also contain information, which is however ignored with that approach. An interesting issue is, therefore, if for economic analyses such as the one at hand, there is an added value by using the whole data set with the imputed missing values compared to using the usually small percentage of complete data (baseline). Also, it is interesting to see how different algorithms affect that result. The imputation of the missing data is done using unsupervised learning. Out of the numerous unsupervised learning approaches, the most common ones, such as clustering, principal component analysis, or neural networks techniques are applied. By training the model iteratively on the imputed data and, thereby, including the information of all data into the model, the distortion of the first training set—the complete data—vanishes. In a next step, the performances of the algorithms are measured. This is done by randomly creating missing values in subsets of the data, estimating those values with the relevant algorithms and several parameter combinations, and comparing the estimates to the actual data. After having found the optimal parameter set for each algorithm, the missing values are being imputed. Using the resulting data sets, the next step is to estimate the willingness to pay for real estate. This is done by fitting price distributions for real estate properties with certain characteristics, such as the region or the number of rooms. Based on these distributions, survival functions are computed to obtain the functional relationship between characteristics and selling probabilities. Comparing the survival functions shows that estimates which are based on imputed data sets do not differ significantly from each other; however, the demand estimate that is derived from the baseline data does. This indicates that the baseline data set does not include all available information and is therefore not representative for the entire sample. Also, demand estimates derived from the whole data set are much more accurate than the baseline estimation. Thus, in order to obtain optimal results, it is important to make use of all available data, even though it involves additional procedures such as data imputation.

Keywords: demand estimate, missing-data imputation, real estate, unsupervised learning

Procedia PDF Downloads 290

Authors: Maxence Queyrel, Edi Prifti, Alexandre Templier, Jean-Daniel Zucker

Abstract:

Analysis of the human microbiome using metagenomic sequencing data has demonstrated high ability in discriminating various human diseases. Raw metagenomic sequencing data require multiple complex and computationally heavy bioinformatics steps prior to data analysis. Such data contain millions of short sequences read from the fragmented DNA sequences and stored as fastq files. Conventional processing pipelines consist in multiple steps including quality control, filtering, alignment of sequences against genomic catalogs (genes, species, taxonomic levels, functional pathways, etc.). These pipelines are complex to use, time consuming and rely on a large number of parameters that often provide variability and impact the estimation of the microbiome elements. Training Deep Neural Networks directly from raw sequencing data is a promising approach to bypass some of the challenges associated with mainstream bioinformatics pipelines. Most of these methods use the concept of word and sentence embeddings that create a meaningful and numerical representation of DNA sequences, while extracting features and reducing the dimensionality of the data. In this paper we present an end-to-end approach that classifies patients into disease groups directly from raw metagenomic reads: metagenome2vec. This approach is composed of four steps (i) generating a vocabulary of k-mers and learning their numerical embeddings; (ii) learning DNA sequence (read) embeddings; (iii) identifying the genome from which the sequence is most likely to come and (iv) training a multiple instance learning classifier which predicts the phenotype based on the vector representation of the raw data. An attention mechanism is applied in the network so that the model can be interpreted, assigning a weight to the influence of the prediction for each genome. Using two public real-life data-sets as well a simulated one, we demonstrated that this original approach reaches high performance, comparable with the state-of-the-art methods applied directly on processed data though mainstream bioinformatics workflows. These results are encouraging for this proof of concept work. We believe that with further dedication, the DNN models have the potential to surpass mainstream bioinformatics workflows in disease classification tasks.

Keywords: deep learning, disease prediction, end-to-end machine learning, metagenomics, multiple instance learning, precision medicine

Procedia PDF Downloads 126