A Wearable Multi-sensor System for Classification of Multiple System Atrophy and Parkinson's Disease

Multiple system atrophy (MSA) is an atypical parkinsonism disorder with faster progression and clinical symptoms similar to Parkinson’s disease (PD). Thus, it is critical to discriminate the diseases as early as possible to provide better therapies for patients and gain the maximum profits. Although some methods, such as positron emission tomography and cerebrospinal fluid, have good performance in clinical practice, those methods are limited since they would increase the body burden and bring extra cost to patients. Recently, significant differences have been proven on spatiotemporal gait features between MSA and PD, however, to the best of our knowledge, there remains no research on making differential diagnosis between MSA and PD by gait analysis. Therefore, in this work, we design a wearable multi-sensor system based on inertial sensors to collect gait information from MSA and PD patients, respectively, and analyze their gait information to make differential diagnosis between the two above diseases with similar symptoms. We evaluated the proposed system on a total 10 MSA and 21 PD patients. As a result, the performance of proposed system reached 89.1% sensitivity, 89.1% specificity and 89.4% accuracy for the classification between MSA and PD.