Visualization of motor symptoms related to Parkinson's disease using wearable sensors

Abstract

Parkinson Disease (PD) is the second most common neurological disease after Alzheimer. There is a need for long term monitoring to determine with higher accuracy the stage of the disease and regulate the levodopa treatment. Current wearable technology can achieve this monitorization of the patient’s daily life. Motor symptoms of the disease are the most evident and thus, the easiest to target and to relate to the stage of the disease. They are also the ones that suppose the highest impediment for the patients to perform their daily living activities. In this study, motor symptoms are assessed via pressure sensitive insole and gyroscopic sensors placed on the wrists. Eight subjects were analyzed, four controls and four with different stages of PD, performing a 20-step walking test. Pressure sensitive insole showed the transfer of force in the foot during the gait cycle. These signals showed the level of pronation and supination of each step. The force applied against the ground was reduced in subjects with PD, and specially seen in the toe-off phase which translate in a reduction in the ankle force. There was no apparent change in the step time in any of the signals. Gyroscopic data evaluation consisted in time domain, frequency domain and spectrogram analysis and comparing the Root Mean Squared (RMS) value and entropy of the signals with the stages of the disease to see any correlations. These procedures were made with the signals measured in the three axes and with the calculated angular velocity vector module. The analysis showed that the tremor could be visualized and the effects of bradykinesia were visible in the signals while walking. RMS and Entropy values didn’t show significance correlation between the stages of PD and their values with the exception of the RMS values of the signals in the Y axis and of the vector module. Tremors appeared in the frequency domain in the form of peaks at 5 Hz that were constant through the test, as shown by the spectrograms. The frequency domain of the vector module had the same peaks as the rest of the signals but at the double of their frequency. Since all the signals correspond to a different person from simple tests, there was no way of assessing the effects of the different stages of the disease in the same individual over time. Wearable technology supposes a good viable solution to the problem of long-term daily monitoring for patients with PD. Suunto Movesense ® gyroscope sensors and Smart insole Forciot ® suppose a good of non-invasive monitoring technology that can provide long term daily data with minimal discomfort while assessing motor dysfunctions that alter the movements of a patient.