Multiple sclerosis (MS) is a prevalent neurological condition affecting the central nervous system and stands as the primary cause of non-traumatic disability among young adults. Traditional diagnostic methods for MS involve clinical laboratory tests and neuroimaging studies. However, the sporadic nature of clinic visits underscores the necessity for remote and frequent monitoring approaches. Such methods facilitate early detection, prompt treatment initiation, and the mitigation of disease progression. This study delves into the exploration of reliable, clinically pertinent features sourced from mobile and wearable technologies. These features aim to differentiate individuals with MS from healthy counterparts, gauge MS-related disability, and assess fatigue levels. By formalizing clinical insights and deriving behavioral indicators, this research endeavors to characterize MS more comprehensively. Evaluation of our methodology employs a dataset comprising 55 individuals with MS and 24 healthy controls, spanning 489 days of observation in natural environments. This dataset incorporates wearable sensor data (e.g., heart rate), smartphone metrics (e.g., phone usage patterns), patient health records (e.g., MS subtype), and self-reported data (e.g., fatigue levels via validated questionnaires). Our findings underscore the viability of leveraging mobile and wearable sensor data for MS monitoring, offering avenues for continuous observation in real-world settings. This approach not only aids in treatment efficacy assessment and disease management but also facilitates participant selection for clinical trials.
Source : Nature.com
