Parkinson’s disease freezing of gait (FOG) symptom detection using machine learning from wearable sensor data

Abstract

Freezing of gait (FoG) is a special symptom found in patients with Parkinson’s disease (PD). Patients who have FoG abruptly lose the capacity to walk as they normally would. Accelerometers worn by patients can record movement data during these episodes, and machine learning algorithms may be able to categorize this information. Thus, the combination may be able to identify FoG in real time. In order to identify FoG events in accelerometer data, we introduce transformer encoder-Bi-LSTM fusion and transformer encoder-GRU fusion models in this study. The model’s capability to differentiate between FoG episodes and normal movement was used to evaluate its performance, and on the Kaggle freezing of gait dataset, the proposed transformer encoder-Bi-LSTM fusion model produced better results with 52.06% compared to combination of transformer encoder and GRU with 49.38% in respect of mean average precision. The findings highlight how Deep Learning-based approaches may progress the field of FoG identification and help PD patients receive better treatments and management plans.