| Summary: | Parkinson's disease (PD) is a debilitating neurodegenerative disorder. Early detection of the disease is important for effective treatment and can improve patients' quality of life. In order to achieve early detection, a population level screening tool is required. This paper introduces a prototype which aims to capture early signs of PD symptoms from users' Chinese handwriting on mobile screens. It can be seamlessly infused into everyday use.
The prototype collects rich features of handwriting in the forms of images and time series data to support pattern analytics research that can assess users' PD risk. A preliminary model based on Convolutional Neural Network and Long Short-Term Memory is built to differentiate between the handwriting of elderly above 60 years old and adults below 30 years old as a proof of concept. 10 subjects were involved in the application testing of writing 50 Chinese characters, with 5 elderly and 5 young adults. An accuracy of $70.8\%$ is achieved from the best model, with a precision of $66.7\%$ at identifying handwriting by elderly. The result could be improved by collecting more samples to reduce overfitting.
In the future, collaborations with healthcare institution for actual data collection with Parkinson's patients can allow more accurate depiction of their medical condition.
|
|---|