Bioelectrics
Elham Dehghanpur Deharab; Peyvand Ghaderyan
Volume 15, Issue 4 , March 2022, , Pages 279-287
Abstract
Parkinson's disease (PD) is one of the most common types of dementia associated with motor impairments and affected performance of motor skills such as writing. Brain imaging techniques are the common methods used to diagnose PD, which are expensive or invasive, and their accuracy depends on the experience ...
Read More
Parkinson's disease (PD) is one of the most common types of dementia associated with motor impairments and affected performance of motor skills such as writing. Brain imaging techniques are the common methods used to diagnose PD, which are expensive or invasive, and their accuracy depends on the experience and the skill of the physician. Therefore, the development of an automated, low cost, and reliable diagnostic system is desirable for researchers. In this study, a handwriting signal including cognitive and motor-perceptual components has been used as a non-invasive, cost effective and reliable characteristic in identifying PD-related cognitive and motor dysfunctions. For this purpose, the matching pursuit algorithm with high time-frequency resolution has been employed to decompose X-Y coordinates. It provides a sparse representation of the handwriting signals and quantifies the basic information about the local changes in the handwriting signals. The proposed method is evaluated on a database with 31 healthy samples and 29 Parkinson's samples using the support vector machine classifier and obtained results yields an average accuracy rate of 90%, sensitivity rate of 91.59% and specificity rate of 90%. Comparing different writing tasks has also demonstrated superior performance of writing an entire sentence for PD detection.
