{"title":"Comparison of Multivariate Adaptive Regression Splines and Random Forest Regression in Predicting Forced Expiratory Volume in One Second","authors":"P. V. Pramila, V. Mahesh","volume":100,"journal":"International Journal of Bioengineering and Life Sciences","pagesStart":338,"pagesEnd":343,"ISSN":"1307-6892","URL":"https:\/\/publications.waset.org\/pdf\/10000957","abstract":"
Pulmonary Function Tests are important non-invasive
\r\ndiagnostic tests to assess respiratory impairments and provides
\r\nquantifiable measures of lung function. Spirometry is the most
\r\nfrequently used measure of lung function and plays an essential role
\r\nin the diagnosis and management of pulmonary diseases. However,
\r\nthe test requires considerable patient effort and cooperation,
\r\nmarkedly related to the age of patients resulting in incomplete data
\r\nsets. This paper presents, a nonlinear model built using Multivariate
\r\nadaptive regression splines and Random forest regression model to
\r\npredict the missing spirometric features. Random forest based feature
\r\nselection is used to enhance both the generalization capability and the
\r\nmodel interpretability. In the present study, flow-volume data are
\r\nrecorded for N= 198 subjects. The ranked order of feature importance
\r\nindex calculated by the random forests model shows that the
\r\nspirometric features FVC, FEF25, PEF, FEF25-75, FEF50 and the
\r\ndemographic parameter height are the important descriptors. A
\r\ncomparison of performance assessment of both models prove that, the
\r\nprediction ability of MARS with the `top two ranked features namely
\r\nthe FVC and FEF25 is higher, yielding a model fit of R2= 0.96 and
\r\nR2= 0.99 for normal and abnormal subjects. The Root Mean Square
\r\nError analysis of the RF model and the MARS model also shows that
\r\nthe latter is capable of predicting the missing values of FEV1 with a
\r\nnotably lower error value of 0.0191 (normal subjects) and 0.0106
\r\n(abnormal subjects) with the aforementioned input features. It is
\r\nconcluded that combining feature selection with a prediction model
\r\nprovides a minimum subset of predominant features to train the
\r\nmodel, as well as yielding better prediction performance. This
\r\nanalysis can assist clinicians with a intelligence support system in the
\r\nmedical diagnosis and improvement of clinical care.<\/p>\r\n","references":"[1] Daniel C Ginnan and Jonathon Dean Truwit, \u201cClinical review:\r\nRespiratory mechanics in spontaneous and assisted ventilation,\u201d Critical\r\nCare, vol. 9, no.5, pp. 472\u2013484, 2005.\r\n[2] R. L. Mulder, N. M. Thonissen, J. H. H. Vander Pal, P. Bresser, W.\r\nHanselaar, C. C. E. Koning, F. Oldenburger, H. A. Heij, H. N. Caron,\r\n\u201cPulmonary function impairment measured by pulmonary function tests\r\nin lon g-term survivors of childhood cancer,\u201d Thorax, vol. 66, pp. 1065-\r\n1071, 2011.\r\n[3] A. Mythili, C. M. Sujatha , S. Srinivasan and S. Ramakrishnan,\r\n\u201cPrediction Of Forced Expiratory Volume In Spirometric Pulmonary\r\nFunction Test Using Adaptive Neuro Fuzzy Inference System,\u201d\r\nBiomedical Sciences Instrumentation, vol. 48, pp.508-15, 2012.\r\n[4] D. Ozerkis-Antin, J. Evans, A. Rubinowitz, R.J. Horner, R.A. Matthay,\r\n\u201cPulmonary manifestations of rheumatoid arthritis,\u201d Clinical Chest\r\nMedicine, vol.31, no.3, pp. 451-78, 2010.\r\n[5] Thomas A Barnes, Len Fromer, \u201cSpirometry use: detection of chronic\r\nobstructive pulmonary disease in the primary care setting,\u201d International\r\nJournal of CODP, 2011.\r\n[6] R.E. Dales, K.L. Vandemheen, J. Clinch, et al. \u201cSpirometry in the\r\nprimary care setting: influence on clinical diagnosis and management of\r\nairflow obstruction,\u201d Journal of Chest, vol.128, no. 4, pp. 2443\u20132447,\r\n2005.\r\n[7] N. Chavannes, T. Schermer, R. Akkermans, et al. \u201cImpact of spirometry\r\non GPs\u2019 diagnostic differentiation and decision-making,\u201d Respiratory\r\nMedicine, vol.98, no.11, pp.1124\u20131130, 2004.\r\n[8] R.P.Young, R. Hopkins, T.E. Eaton, \u201cForced expiratory volume in one\r\nsecond: not just a lung function test but a marker of premature death\r\nfrom all causes,\u201d European Respiratory Journal, vol. 30, no.4, pp.616\u2013\r\n622, 2007.\r\n[9] \u201cStandards for the diagnosis and care of patients with chronic\r\nobstructive pulmonary disease,\u201d American Thoracic Society, American\r\nJournal of Respiratory and Critical Care Med, vol.152, pp.77-121, 1995.\r\n[10] D.C. Richter , J.R. Joubert , H. Nell, M.M. Schuurmans, E.M. Irusen,\r\n\u201cDiagnostic value of post-bronchodilator pulmonary function testing to\r\ndistinguish between stable, moderate to severe COPD and asthma,\r\nInternational journal of chronic obstructive pulmonary disorder, vol. 3,\r\nno.4, pp. 693-699, 2008.\r\n[11] Jeffrey M. Haynes, \u201cPulmonary Function Test Quality in the Elderly: A\r\nComparison with Younger Adults,\u201d Respiratory care, vol.59, no.1, jan\r\n2014.\r\n[12] American Thoracic Society, Standardization of spirometry: a summary\r\nof recommendations from the American Thoracic Society. 1987 update,\r\nAnn Intern Med, vol.108, pp. 217\u2013220, 1988.\r\n[13] V. Bellia, R. Pistelli, F. Catalano, R. Antonelli-Incalzi, V. Grassi, G.\r\nMeillo, et al . \u201cQuality control of spirometry in the elderly: the SARA\r\nstudy,\u201d Am J Respir Crit Care Med, vol. 161, no.4, pp. 1094-1100, 2000.\r\n[14] L. Pezzoli, G. Giardini, S. Consonni , I. Dallera, C. Bilotta, G. Ferrario\r\nG et al. \u201cQuality of sprirometric performance in older people. Age\r\nAgeing,\u201d vol. 32, no. 1, pp. 43-46, 2003.\r\n[15] Xu, Ruo, \u201cImprovements to random forest methodology,\u201d Graduate\r\nThesis and Dissertations, Paper 13052, 2013.\r\n[16] Mark R. Segal, \u201cMachine Learning Benchmarks and Random Forest\r\nRegression,\u201d Kluwer Academic Publishers, 2003.\r\n[17] L. Breiman, \u201cRandom forests,\u201d Machine Learning, vol. 45, 2001, pp. 5\u2013\r\n32.\r\n[18] Anne-Laure Boulesteix, Silke Janitza, Jochen Kruppa, Inke R. Konig,\r\n\u201cOverview of Random Forest Methodology and Practical Guidance with\r\nEmphasis on Computational Biology and Bioinformatics,\u201d available at:\r\nhttp:\/\/epub.ub.uni-muenchen.de\/13766\/1\/TR.pdf [19] M. Hilario, A. Kalousis, C. Pellegrini, M. Muller, \u201cProcessing and\r\nclassification of protein mass spectra, Mass Spectrom Rev, vol.25, pp.\r\n409-449.\r\n[20] Akin \u00d6z\u00e7ift, \u201cRandom forests ensemble classifier trained with data\r\nresampling strategy to improve cardiac arrhythmia diagnosis,\u201d\r\nComputers in Biology and Medicine, vol.41, no.5, pp.265-271, 2011.\r\n[21] Benjamin A Goldstein, Alan E Hubbard, Adele Cutler and Lisa F\r\nBarcellos, \u201cAn application of Random Forests to a genome-wide\r\nassociation dataset: Methodological considerations & new findings,\u201d\r\nBMC genetics, vol.11, no.49, 2010.\r\n[22] Nahit Emanet, Halil R \u00d6z, Nazan Bayram and Dursun Delen, \u201cA\r\ncomparative analysis of machine learning methods for classification type\r\ndecision problems in healthcare,\u201d Decision Analytics, vol.1, no.6, pp.1-\r\n20, 2014.\r\n[23] K. J. Archer and R.V. Kimes, \u201cEmpirical characterization of random\r\nforest variable importance measures,\u201d Computational Statistics and Data\r\nAnalysis, vol. 52, pp. 2249-2260.\r\n[24] J.H. Friedman, Multivariate adaptive regression splines, Ann. Stat., vol.\r\n19, pp. 1\u2013141, 1991.\r\n[25] Dani Guzm\u00e1n, Francisco Javier de Cos Juez, Fernando S\u00e1nchez\r\nLasheras, Richard loop adaptiev Myers and Laura Young, \u201cDeformable\r\nmirror model for open- optics using multivariate adaptive regression\r\nsplines,\u201d Optics Express, vol.18, no.7, pp. 6492 \u2013 6505, 2013.\r\n[26] P. A. W. Lewis and J. G. Stevens, \u201cNonlinear modeling of time series\r\nusing multivariate adaptive regression splines (mars),\u201d Journal of the\r\nAmerican Statistical Association, vol. 86, no. 416, pp. 864-877, 1991.\r\n[27] Peter C. Austin, \u201cA comparison of regression trees, logistic regression,\r\ngeneralized additive models, and multivariate adaptive regression splines\r\nfor predicting AMI mortality,\u201d Statistics in Medicine, vol. 26, pp. 2937\u2013\r\n2957,","publisher":"World Academy of Science, Engineering and Technology","index":"Open Science Index 100, 2015"}