TY - GEN
T1 - Automatic detection of Spasticity from flexible wearable sensors
AU - Lonini, Luca
AU - Shawen, Nicholas
AU - Ghaffari, Roozbeh
AU - Rogers, John
AU - Jayarman, Arun
N1 - Funding Information:
This work was funded by the Max Näder Lab.
Publisher Copyright:
Copyright © 2017 ACM.
PY - 2017/9/11
Y1 - 2017/9/11
N2 - Spasticity is a condition that impairs voluntary muscle movements and physically debilitates persons across several neurological disorders, including stroke, multiple sclerosis and cerebral palsy. Assessing the progression of spasticity during clinical interventions and at home is key to rehabilitation efficacy and care management. Here we present electromyography (EMG) and motion data using skin-mounted, flexible and wireless sensors in a cohort of 13 individuals with stroke. We compute a set of 15 features from the EMG data and use machine learning to infer whether spasticity is present during movements of the knee and ankle joints. Using a Linear Discriminant Analysis (LDA) classifier, we show that our approach successfully discriminates voluntary contractions from spastic muscle contractions (AUC=0.94). These results show that continuous and non-invasive monitoring of spasticity symptoms could be applied to optimize and personalize rehabilitation regimens.
AB - Spasticity is a condition that impairs voluntary muscle movements and physically debilitates persons across several neurological disorders, including stroke, multiple sclerosis and cerebral palsy. Assessing the progression of spasticity during clinical interventions and at home is key to rehabilitation efficacy and care management. Here we present electromyography (EMG) and motion data using skin-mounted, flexible and wireless sensors in a cohort of 13 individuals with stroke. We compute a set of 15 features from the EMG data and use machine learning to infer whether spasticity is present during movements of the knee and ankle joints. Using a Linear Discriminant Analysis (LDA) classifier, we show that our approach successfully discriminates voluntary contractions from spastic muscle contractions (AUC=0.94). These results show that continuous and non-invasive monitoring of spasticity symptoms could be applied to optimize and personalize rehabilitation regimens.
KW - Flexible electronics
KW - Machine learning
KW - Rehabilitation
UR - http://www.scopus.com/inward/record.url?scp=85030859453&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85030859453&partnerID=8YFLogxK
U2 - 10.1145/3123024.3123098
DO - 10.1145/3123024.3123098
M3 - Conference contribution
AN - SCOPUS:85030859453
T3 - UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers
SP - 133
EP - 136
BT - UbiComp/ISWC 2017 - Adjunct Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers
PB - Association for Computing Machinery, Inc
T2 - 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and ACM International Symposium on Wearable Computers, UbiComp/ISWC 2017
Y2 - 11 September 2017 through 15 September 2017
ER -