AI Techniques To Automatically Detect Standardized Functional Test Patterns From Wearable Sensor Data

Wearable devices are often utilized for the diagnosis, treatment, and rehabilitation of various diseases. Data acquired from them may be used to monitor patient recovery during rehabilitation or recovery. Wearable devices allow for the collection of data from various standardized functional assessments that quantify activities of daily living (ADL). The volume of data collected by the sensors from detailed recordings of long-term movement data can become enormous, making it challenging to manually extract movement information with any degree of certainty. The goal of this research was to create an automated system capable of extracting standardized functional test (SFT) patterns from wearable devices worn in an ambulatory environment. The datasets used in this study contained movement data from 40 people living with axial spondylo arthritis. Each participant in this study completed a series of SFTs. The first session was completed in a clinical setting, and the second session at each participant's home. An artificial intelligence (AI) system was developed to automatically extract SFT movement patterns from the long-term datasets. The resultant model demonstrated an accuracy of 97.37%.