Method of Early Diagnosis of Abnormal Movements of Patients with Tremor Symptoms Using Augmented Reality

The work aims to increase the accuracy and earlier diagnosis of tremor, considering it as a key symptom of Parkinson's disease and a manifestation of neurological disorders caused by injuries. Unlike traditional methods, the proposed approach eliminates the presence of a compensatory "anchor", increasing sensitivity to the slightest manifestations of tremor. The efficiency of the method was experimentally confirmed on 50 healthy people, 6 with Parkinson's disease, and 75 with traumatic brain injuries resulting from military operations. The method was compared with the classical assessment of drawing geometric figures on paper. A comparison showed that all patients with Parkinson’s disease passed the paper test, but the most difficult of them could not pass the AR test without a fulcrum. For wounded servicemen, AR showed a visible advantage, though significantly more tests are required to use a deep classifier. Studies using a basic classifier have shown that the proposed method, which utilizes augmented reality as a medium, is more sensitive than the standard one, confirming its promise for early neurodiagnostic systems.