- Browse by Author
Browsing by Author "Suresh, Shruthi"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Automatic Detection and Characterization of Autonomic Dysreflexia Using Multi-Modal Non-Invasive Sensing and Neural Networks(Mary Ann Liebert, 2022-11-10) Suresh, Shruthi; Everett, Thomas H., IV; Shi, Riyi; Duerstock, Bradley S.; Medicine, School of MedicineAutonomic dysreflexia (AD) frequently occurs in persons with spinal cord injuries (SCIs) above the T6 level triggered by different stimuli below the level of injury. If improperly managed, AD can have severe clinical consequences, even possibly leading to death. Existing techniques for AD detection are time-consuming, obtrusive, lack automated detection capabilities, and have low temporal resolution. Therefore, a non-invasive, multi-modal wearable diagnostic tool was developed to quantitatively characterize and distinguish unique signatures of AD. Electrocardiography and novel skin nerve activity (skNA) sensors with neural networks were used to detect temporal changes in the sympathetic and vagal systems in rats with SCI. Clinically established metrics of AD were used to verify the onset of AD. Five physiological features reflecting different metrics of sympathetic and vagal activity were used to characterize signatures of AD. An increase in sympathetic activity, followed by a lagged increase in vagal activity during the onset of AD, was observed after inducing AD. This unique signature response was used to train a neural network to detect the onset of AD with an accuracy of 93.4%. The model also had a 79% accuracy in distinguishing between sympathetic hyperactivity reactions attributable to different sympathetic stressors above and below the level of injury. These neural networks have not been used in previous work to detect the onset of AD. The system could serve as a complementary non-invasive tool to the clinically accepted gold standard, allowing an improved management of AD in persons with SCI.Item Feature Selection Techniques for a Machine Learning Model to Detect Autonomic Dysreflexia(Frontiers Media, 2022-08-10) Suresh, Shruthi; Newton, David T.; Everett, Thomas H., IV; Lin, Guang; Duerstock, Bradley S.; Medicine, School of MedicineFeature 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.