Parameter estimate from accelerometer and gyroscope for characterization of wrist tremor in individuals with parkinson’s disease

Abstract

Researchers have been working on the quantification of tremor to assess the severity of neurodegenerative diseases that present it as a pathological symptom. Usually, clinical rating scales are employed for the qualitative assessment of tremor. This research presents quantitative results, estimated by the root mean square (RMS) of signals estimated from accelerometers and gyroscopes, to quantify the degree of severity of tremor in individuals with Parkinson’s disease (PD). In total, 27 individuals participated in the study, being 15 with PD (GI) and 12 healthy (GII). Two set of inertial sensors capable of jointly measuring acceleration and angular speed were positioned respectively on the back of the hand and forearm of the participants. The detected signals were digitized and annotated for the identification of the beginning and end of the region of interest. Signals were acquired while the wrist was maintained at rest for 15 s, without and with a load of 92 g on the back of the hand. Three data collections were performed with 60-s intervals. From the RMS it was defined an index (In) that showed to be sensitive to the variation of the amplitude of tremor. The index In showed that PD tremor is larger than that for the healthy participants, and that the addition of load significantly reduces the tremor in PD. This research is relevant because it quantifies the tremor by means of an index that takes into account both patterns of normality and severity of tremor.