Parkinson’s Disease Detection Using Spiral Images (Hand Drawings)

Parkinson’s disease (PD), or simply Parkinson’s, is a long-term degenerative disorder of the central nervous system that mainly affects the motor system. The symptoms usually emerge slowly, and as the disease worsens, non-motor symptoms become more common. The most obvious early symptoms are tremor, rigidity, slowness of movement, and difficulty with walking.

Parkinson’s Disease can cause slow motor movements, depression, anxiety, sleep and sensory system disorders, behaviour changes etc. Environmental factors and Genetic Inheritance are amongst the major contributing factors for Parkinson’s disease.

Spiral drawing is a skilled and complex coordinated motor activity. Therefore, it is treated as a sensitive motor assessment and a preliminary test for early symptoms of Parkinson’s disease.

This article presents a solution for detecting Parkinson’s disease using Spiral Drawings and Convolutional Neural Networks (CNN). The Cainvas Platform is used for implementation, which provides seamless execution of python notebooks for building AI systems that can eventually be deployed on edge (i.e. an embedded system such as compact MCUs).
The notebook can be found here.

The flow of the article is as follows —

Description of the Problem Statement
The Dataset
Data Augmentation
CNN Model Architecture
Training the Model
Performance of Model
Testing models on Images
Conclusion

Description of the Problem Statement
The project aims at presenting a solution for Parkinson’s disease detection using Spiral Drawings and CNN. The main idea behind the implementation is to classify a person as Healthy or having Parkinson’s disease by looking at the Spiral Drawing made by the person.

The Spiral Drawing created by a healthy person will look almost similar to a standard spiral shape. However, a spiral drawn by a person with Parkinson’s disease will highly deviate from a perfect spiral shape and look distorted due to slow motor movements and decreased coordination between hand and brain.

The Dataset
The dataset used here is the Parkinson’s Drawing Dataset present on Kaggle. The dataset can be accessed through this link. The dataset contains Spiral and Waves drawings made by healthy people and Parkinson’s disease infected people.

In this article, only Spiral Drawings are used for classification. The dataset already contains the Train Set and the Test Set, so no need to manually split the dataset.