Feature Selection Techniques for a Machine Learning Model to Detect Autonomic Dysreflexia
Abstract
Feature selection plays a crucial role in the development of machine learning algorithms. Understanding the impact of the features on a model, and their physiological relevance can improve the performance. This is particularly helpful in the healthcare domain wherein disease states need to be identified with relatively small quantities of data. Autonomic Dysreflexia (AD) is one such example, wherein mismanagement of this neurological condition could lead to severe consequences for individuals with spinal cord injuries. We explore different methods of feature selection needed to improve the performance of a machine learning model in the detection of the onset of AD. We present different techniques used as well as the ideal metrics using a dataset of thirty-six features extracted from electrocardiograms, skin nerve activity, blood pressure and temperature. The best performing algorithm was a 5-layer neural network with five relevant features, which resulted in 93.4% accuracy in the detection of AD. The techniques in this paper can be applied to a myriad of healthcare datasets allowing forays into deeper exploration and improved machine learning model development. Through critical feature selection, it is possible to design better machine learning algorithms for detection of niche disease states using smaller datasets.
- Authors:
-
- Purdue University, West Lafayette, IN (United States)
- Indiana University School of Medicine, Indianapolis IN (United States)
- Publication Date:
- Research Org.:
- Purdue Univ., West Lafayette, IN (United States)
- Sponsoring Org.:
- USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR); Department of Defense (DOD); Indiana State Department of Health; National Institutes of Health (NIH); Krannert Institute of Cardiology; National Science Foundation (NSF); Army Research Office (ARO)
- OSTI Identifier:
- 1981089
- Grant/Contract Number:
- SC0021142; SC190164; R42DA043391; OT2OD028183-01; DMS-1555072; DMS-1736364; CMMI-1634832; CMMI-1560834; 382247; W911NF-15-1-0562
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Frontiers in Neuroinformatics
- Additional Journal Information:
- Journal Volume: 16; Journal ID: ISSN 1662-5196
- Publisher:
- Frontiers Media S.A.
- Country of Publication:
- United States
- Language:
- English
- Subject:
- spinal cord injuries; machine learning; feature selection; electrocardiography; healthcare
Citation Formats
Suresh, Shruthi, Newton, David T., Everett, Thomas H., Lin, Guang, and Duerstock, Bradley S. Feature Selection Techniques for a Machine Learning Model to Detect Autonomic Dysreflexia. United States: N. p., 2022.
Web. doi:10.3389/fninf.2022.901428.
Suresh, Shruthi, Newton, David T., Everett, Thomas H., Lin, Guang, & Duerstock, Bradley S. Feature Selection Techniques for a Machine Learning Model to Detect Autonomic Dysreflexia. United States. https://doi.org/10.3389/fninf.2022.901428
Suresh, Shruthi, Newton, David T., Everett, Thomas H., Lin, Guang, and Duerstock, Bradley S. Wed .
"Feature Selection Techniques for a Machine Learning Model to Detect Autonomic Dysreflexia". United States. https://doi.org/10.3389/fninf.2022.901428. https://www.osti.gov/servlets/purl/1981089.
@article{osti_1981089,
title = {Feature Selection Techniques for a Machine Learning Model to Detect Autonomic Dysreflexia},
author = {Suresh, Shruthi and Newton, David T. and Everett, Thomas H. and Lin, Guang and Duerstock, Bradley S.},
abstractNote = {Feature selection plays a crucial role in the development of machine learning algorithms. Understanding the impact of the features on a model, and their physiological relevance can improve the performance. This is particularly helpful in the healthcare domain wherein disease states need to be identified with relatively small quantities of data. Autonomic Dysreflexia (AD) is one such example, wherein mismanagement of this neurological condition could lead to severe consequences for individuals with spinal cord injuries. We explore different methods of feature selection needed to improve the performance of a machine learning model in the detection of the onset of AD. We present different techniques used as well as the ideal metrics using a dataset of thirty-six features extracted from electrocardiograms, skin nerve activity, blood pressure and temperature. The best performing algorithm was a 5-layer neural network with five relevant features, which resulted in 93.4% accuracy in the detection of AD. The techniques in this paper can be applied to a myriad of healthcare datasets allowing forays into deeper exploration and improved machine learning model development. Through critical feature selection, it is possible to design better machine learning algorithms for detection of niche disease states using smaller datasets.},
doi = {10.3389/fninf.2022.901428},
journal = {Frontiers in Neuroinformatics},
number = ,
volume = 16,
place = {United States},
year = {Wed Aug 10 00:00:00 EDT 2022},
month = {Wed Aug 10 00:00:00 EDT 2022}
}
Works referenced in this record:
Integration of Early Physiological Responses Predicts Later Illness Severity in Preterm Infants
journal, September 2010
- Saria, Suchi; Rajani, Anand K.; Gould, Jeffrey
- Science Translational Medicine, Vol. 2, Issue 48
Automated feature selection for obstructive sleep apnea syndrome diagnosis
journal, January 2020
- Wosiak, Agnieszka; Kowalski, Rafał
- Procedia Computer Science, Vol. 176
An Introduction to Machine Learning
journal, March 2020
- Badillo, Solveig; Banfai, Balazs; Birzele, Fabian
- Clinical Pharmacology & Therapeutics, Vol. 107, Issue 4
Artificial intelligence and machine learning to fight COVID-19
journal, April 2020
- Alimadadi, Ahmad; Aryal, Sachin; Manandhar, Ishan
- Physiological Genomics, Vol. 52, Issue 4
Estimating blood pressure trends and the nocturnal dip from photoplethysmography
journal, February 2019
- Radha, Mustafa; de Groot, Koen; Rajani, Nikita
- Physiological Measurement, Vol. 40, Issue 2
Application of classification techniques on development an early-warning system for chronic illnesses
journal, August 2012
- Jen, Chih-Hung; Wang, Chien-Chih; Jiang, Bernard C.
- Expert Systems with Applications, Vol. 39, Issue 10
Using skin sympathetic nerve activity to estimate stellate ganglion nerve activity in dogs
journal, June 2015
- Jiang, Zhaolei; Zhao, Ye; Doytchinova, Anisiia
- Heart Rhythm, Vol. 12, Issue 6
Machine Learning for Healthcare: On the Verge of a Major Shift in Healthcare Epidemiology
journal, August 2017
- Wiens, Jenna; Shenoy, Erica S.
- Clinical Infectious Diseases, Vol. 66, Issue 1
Detection of ECG Arrhythmia using a differential expert system approach based on principal component analysis and least square support vector machine
journal, March 2007
- Polat, Kemal; Güneş, Salih
- Applied Mathematics and Computation, Vol. 186, Issue 1
Recording sympathetic nerve activity from the skin
journal, October 2017
- Everett, Thomas H.; Doytchinova, Anisiia; Cha, Yong-Mei
- Trends in Cardiovascular Medicine, Vol. 27, Issue 7
Automatic detection of respiratory arrests in OSA patients using PPG and machine learning techniques
journal, October 2016
- Uçar, Muhammed Kürşad; Bozkurt, Mehmet Recep; Bilgin, Cahit
- Neural Computing and Applications, Vol. 28, Issue 10
Automated machine learning: Review of the state-of-the-art and opportunities for healthcare
journal, April 2020
- Waring, Jonathan; Lindvall, Charlotta; Umeton, Renato
- Artificial Intelligence in Medicine, Vol. 104
One-Dimensional Scattering of Fermions on δ-Impurities
journal, August 2019
- Guilarte, Juan Mateos; Munoz-Castaneda, Jose M.; Pirozhenko, Irina
- Frontiers in Physics, Vol. 7
Development of an Algorithm to Perform a Comprehensive Study of Autonomic Dysreflexia in Animals with High Spinal Cord Injury Using a Telemetry Device
journal, July 2016
- Popok, David; West, Christopher; Frias, Barbara
- Journal of Visualized Experiments, Issue 113
Rodent Models of Colorectal Distension
journal, October 2012
- O'Mahony, Siobhain M.; Tramullas, Monica; Fitzgerald, Patrick
- Current Protocols in Neuroscience, Vol. 61, Issue 1
A review of feature selection techniques in bioinformatics
journal, August 2007
- Saeys, Y.; Inza, I.; Larranaga, P.
- Bioinformatics, Vol. 23, Issue 19
Measuring Blood Pressure in Mice using Volume Pressure Recording, a Tail-cuff Method
journal, May 2009
- Daugherty, Alan; Rateri, Debra; Hong, Lu
- Journal of Visualized Experiments, Issue 27
Effects of gabapentin on muscle spasticity and both induced as well as spontaneous autonomic dysreflexia after complete spinal cord injury
journal, January 2012
- Rabchevsky, Alexander G.; Patel, Samir P.; Lyttle, Travis S.
- Frontiers in Physiology, Vol. 3
Disease Prediction by Machine Learning Over Big Data From Healthcare Communities
journal, January 2017
- Chen, Min; Hao, Yixue; Hwang, Kai
- IEEE Access, Vol. 5
Ambulatory Blood Pressure Monitoring in Spinal Cord Injury: Clinical Practicability
journal, May 2014
- Hubli, Michèle; Krassioukov, Andrei V.
- Journal of Neurotrauma, Vol. 31, Issue 9
R-Peak Detection Algorithm for Ecg using Double Difference And RR Interval Processing
journal, January 2012
- Sadhukhan, Deboleena; Mitra, Madhuchhanda
- Procedia Technology, Vol. 4
Dermatologist-level classification of skin cancer with deep neural networks
journal, January 2017
- Esteva, Andre; Kuprel, Brett; Novoa, Roberto A.
- Nature, Vol. 542, Issue 7639
Feature selection and classification systems for chronic disease prediction: A review
journal, November 2018
- Jain, Divya; Singh, Vijendra
- Egyptian Informatics Journal, Vol. 19, Issue 3
Data Mining Techniques for Assisting the Diagnosis of Pressure Ulcer Development in Surgical Patients
journal, April 2011
- Su, Chao-Ton; Wang, Pa-Chun; Chen, Yan-Cheng
- Journal of Medical Systems, Vol. 36, Issue 4
Predicting early psychiatric readmission with natural language processing of narrative discharge summaries
journal, October 2016
- Rumshisky, A.; Ghassemi, M.; Naumann, T.
- Translational Psychiatry, Vol. 6, Issue 10
Continuous and online analysis of heart rate variability
journal, January 2005
- Chan, Hl; Chou, Ws; Chen, Sw
- Journal of Medical Engineering & Technology, Vol. 29, Issue 5
Creating Value In Health Care Through Big Data: Opportunities And Policy Implications
journal, July 2014
- Roski, Joachim; Bo-Linn, George W.; Andrews, Timothy A.
- Health Affairs, Vol. 33, Issue 7
Data Mining in Healthcare – A Review
journal, January 2015
- Jothi, Neesha; Rashid, Nur’Aini Abdul; Husain, Wahidah
- Procedia Computer Science, Vol. 72
Comparison of Baseline Wander Removal Techniques considering the Preservation of ST Changes in the Ischemic ECG: A Simulation Study
journal, January 2017
- Lenis, Gustavo; Pilia, Nicolas; Loewe, Axel
- Computational and Mathematical Methods in Medicine, Vol. 2017