Contactless finger tapping detection at C-Band

The rapid finger tap test is widely used in clinical assessment of dyskinesias in Parkinson's disease. In clinical practice, doctors rely on their clinical experience and use the Parkinson's Disease Uniform Rating Scale to make a brief judgment of symptoms. We propose a novel C-band microwave sensing method to evaluate finger tapping quantitatively and qualitatively in a non-contact way based on wireless channel information (WCI). The phase difference between adjacent antennas is used to calibrate the original random phase. Outlier filtering and smoothing filtering are used to process WCI waveforms. Based on the resulting signal, we define and extract a set of features related to the features described in UPDRS. Finally, the features are input into a support vector machine (SVM) to obtain results for patients with different severity. The results show that the proposed system can achieve an average accuracy of 99%. Compared with the amplitude, the average quantization accuracy of the phase difference on finger tapping is improved by 3%. In the future, the proposed system could assist doctors to quantify the movement disorders of patients, and it is very promising to be a candidate for clinical practice.