The effect of Parkinson's disease progression on RQA chaotic features

Parkinson's disease is a neurodegenerative disease that causes severe movement disorders including bradykinesia, rigidity, and tremors. There is no cure for Parkinson's disease, only the symptoms can be managed. Parkinson's disease is diagnosed using the MDS-UPDRS global grading scale. In this scale, four levels including slight, mild, moderate, and severe levels are defined for the disease. Recurrence plots and RQA features are tools for describing the behavior of chaotic systems and revealing hidden patterns in system dynamics. In this paper, the effect of Parkinson's disease progression on RQA chaotic features is studied. For this purpose, the dataset of the accelerometer mounted on the hand during the finger tapping test was used, which included 67 healthy data, 54 level one data, 66 level two data, 59 level three data, and 14 level four data. After pre-processing, the recurrence plots of the data were drawn and their RQA characteristics were calculated. Patterns of recurrence plots including separate recurrence points, diagonal lines, vertical lines, black squares, and horizontal and vertical white bands were investigated. According to the obtained results, the patterns of recurrence plots had significant differences among different levels of Parkinson's disease. Therefore, RQA features can be used to automatically determine the level of Parkinson's disease.