TY - GEN
T1 - Mean Pressure Gradient Prediction Based on Chest Angular Movements and Heart Rate Variability Parameters
AU - Shokouhmand, Arash
AU - Yang, Chenxi
AU - Aranoff, Nicole D.
AU - Driggin, Elissa
AU - Green, Philip
AU - Tavassolian, Negar
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - This study presents our recent findings on the classification of mean pressure gradient using angular chest movements in aortic stenosis (AS) patients. Currently, the severity of aortic stenosis is measured using ultra-sound echocardiography, which is an expensive technology. The proposed framework motivates the use of low-cost wearable sensors, and is based on feature extraction from gyroscopic readings. The feature space consists of the cardiac timing intervals as well as heart rate variability (HRV) parameters to determine the severity of disease. State-of-the-art machine learning (ML) methods are employed to classify the severity levels into mild, moderate, and severe. The best performance is achieved by the Light Gradient-Boosted Machine (Light GBM) with an F1-score of 94.29% and an accuracy of 94.44%. Additionally, game theory-based analyses are employed to examine the top features along with their average impacts on the severity level. It is demonstrated that the isovolumetric contraction time (IVCT) and isovolumetric relaxation time (IVRT) are the most representative features for AS severity.Clinical Relevance - The proposed framework could be an appropriate low-cost alternative to ultra-sound echocardiography, which is a costly method.
AB - This study presents our recent findings on the classification of mean pressure gradient using angular chest movements in aortic stenosis (AS) patients. Currently, the severity of aortic stenosis is measured using ultra-sound echocardiography, which is an expensive technology. The proposed framework motivates the use of low-cost wearable sensors, and is based on feature extraction from gyroscopic readings. The feature space consists of the cardiac timing intervals as well as heart rate variability (HRV) parameters to determine the severity of disease. State-of-the-art machine learning (ML) methods are employed to classify the severity levels into mild, moderate, and severe. The best performance is achieved by the Light Gradient-Boosted Machine (Light GBM) with an F1-score of 94.29% and an accuracy of 94.44%. Additionally, game theory-based analyses are employed to examine the top features along with their average impacts on the severity level. It is demonstrated that the isovolumetric contraction time (IVCT) and isovolumetric relaxation time (IVRT) are the most representative features for AS severity.Clinical Relevance - The proposed framework could be an appropriate low-cost alternative to ultra-sound echocardiography, which is a costly method.
UR - http://www.scopus.com/inward/record.url?scp=85122504694&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85122504694&partnerID=8YFLogxK
U2 - 10.1109/EMBC46164.2021.9630805
DO - 10.1109/EMBC46164.2021.9630805
M3 - Conference contribution
C2 - 34892754
AN - SCOPUS:85122504694
T3 - Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS
SP - 7170
EP - 7173
BT - 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2021
T2 - 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2021
Y2 - 1 November 2021 through 5 November 2021
ER -