IS in Healthcare
Loading...
Paper Number
1388
Paper Type
Completed
Description
People are living much longer lives due to the advancement of medicine and healthy lifestyles. However, chronic conditions have become an increasing societal issue. Parkinson’s disease (PD), costing $15.5 billion per year for the U.S., is the second most common neurodegenerative disorder in the U.S. To manage PD, health profiling tools project a trajectory for senior citizen’s disease progression, which assist therapies and interventions. With the advancement of sensor technologies, motion sensors have emerged as an effective and efficient approach to collecting motion data from senior citizens. However, few studies have examined motion sensors in health profiling. In this study, we propose a novel deep learning model, the Adaptive Time-aware Convolutional Long Short-Term Memory (ATCLSTM), to accurately assess PD severity with walking experiments over time. We collected a publicly available dataset to test our proposed model. Results show that ATCLSTM significantly and consistently outperforms state-of-the-art benchmark models.
Recommended Citation
Yu, Shuo, "Motion Sensor-Based Health Profiling for Parkinson’s Disease: A Deep Learning Approach" (2021). ICIS 2021 Proceedings. 3.
https://aisel.aisnet.org/icis2021/is_health/is_health/3
Motion Sensor-Based Health Profiling for Parkinson’s Disease: A Deep Learning Approach
People are living much longer lives due to the advancement of medicine and healthy lifestyles. However, chronic conditions have become an increasing societal issue. Parkinson’s disease (PD), costing $15.5 billion per year for the U.S., is the second most common neurodegenerative disorder in the U.S. To manage PD, health profiling tools project a trajectory for senior citizen’s disease progression, which assist therapies and interventions. With the advancement of sensor technologies, motion sensors have emerged as an effective and efficient approach to collecting motion data from senior citizens. However, few studies have examined motion sensors in health profiling. In this study, we propose a novel deep learning model, the Adaptive Time-aware Convolutional Long Short-Term Memory (ATCLSTM), to accurately assess PD severity with walking experiments over time. We collected a publicly available dataset to test our proposed model. Results show that ATCLSTM significantly and consistently outperforms state-of-the-art benchmark models.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
17-Health