Identifying People with Parkinson's Disease using Foot Pressure Data

In this paper, we propose a method to distinguish healthy people from those suffering from Parkinson's disease using foot pressure data, Fast Fourier Transform (FFT), and Principal Component Analysis (PCA). We applied an FFT based on the Hamming method to extract frequency ranges of pressure data from the left and right feet of subjects. We used PCA to reduce the dimensions of features generated by FFT. A neural Network with Weighted Fuzzy Membership functions (NEWFM) was used to distinguish healthy subjects from those with Parkinson's disease. Our method yielded accuracy, specificity and sensitivity values of 75.90%, 61.41%, and 81.09%, respectively.