Asynchronous Assessment of the Relative Intensity of an Online Dance for Parkinson’s Class Using Heart Rate and Video Data
Year: 2026
Type: article
Abstract: Background: Regular participation in physical activity (PA) is positively associated with greater functional independence among adults living with Parkinson’s Disease (PD); however, little is known about the dose of PA that adults receive during dance classes specifically designed for adults with PD. Therefore, the purpose of this study was to evaluate the relative PA intensity of Dance for PD® classes that are conducted online and to preliminarily calibrate a multimodal algorithm for detecting the PA intensity of online, group-based Dance for PD® classes using heart rate and 2D video data. Methods: Adults with a diagnosis of idiopathic PD were recruited as participants; a total of n=15 adults with PD [ages 46 to 84 years; female 93.3% (14)] completed this study. Each participant wore a wireless heart rate (HR) monitor during two ≈60 min group-based, online Dance for PD® classes that were held on separate days, and HR data were recorded asynchronously. A signal processing and unsupervised machine learning pipeline was implemented to align HR data across participants and automatically differentiate group-level PA intensity using HR data. Results: HR responses were found to be stable (ICC: 0.93) across exposures to the same online dance class. The average relative intensity of Dance for PD® classes was 65% HRmax (range 48% to 78%) during the initial Dance for PD® class, and the mean within-subjects difference in the relative PA intensity between classes was 0.67% (2.9) HRmax. The novel pipeline for classifying lower- versus higher-intensity PA segments resulted in a segmentation solution that successfully distinguished respective lesson activities quantitatively and qualitatively using heart rate and video data. Conclusions: An online, group-based Dance for PD® class elicited a moderate PA intensity among middle-aged to older adults living with PD at the group level. A novel multimodal signal analysis pipeline successfully distinguished between lower and higher PA intensity behaviors at the group level using HR data, in addition to providing rich qualitative descriptions about the motor behaviors respectively associated with each PA intensity class using video data.
Source: Sensors
DOI: 10.3390/s26217075
Language: English
Topic: Physical Activity and Health
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,137
OpenAlex ID: W7225644466