LEAP MOTION CONTROLLER BASED HAND TREMOR DETECTION AND ANALYSIS
Habib Ali, Mohammad Samheel, Venkateswara Venkatesan
Manipal Institute of Technology, India
Parkinson’s disease and other neurological conditions often manifest in the form of tremors, especially hand tremors. Tremors can be due to various reasons and based on the type of the tremor, the condition can be diagnosed. Current clinical practice involves visual assessment of the tremor based on various rating scales, which is quite subjective and without any quantification. As a result, there is considerable variability among clinicians in the diagnosis as well. It has been found that the error rate in diagnosing tremors can be as high as 35% in developed countries. Attempts have been made to quantify tremors using accelerometers and Inertial Measurement Units (IMU), laser based systems, video based and magnetic induction based systems. Majority of the work related to tremors used accelerometers or IMU’s. Studies showing the effect of load on tremor have raised concerns regarding change in tremor frequency and amplitude by attaching sensors to the body. In order to avoid any uncertainty in this regard, this paper attempts to use the Leap Motion Controller, which is a small, portable and contactless USB peripheral device which can measure hand position based on IR, to quantify certain frequency domain parameters of tremor. Almost all the existing methods use data from a single finger. This research takes advantage of the Leap Motion Controller’s ability to detect all the five fingers and palm positon. Data from healthy volunteers and volunteers diagnosed with different neurological disorders has been collected and analyzed. The techniques, observations and findings are presented in the paper along with scope for further improvement.