ParkinsonNet

Overview

Millions of people worldwide have been affected by Parkinson's disease (PD), which is characterized by a loss of mobility and, as a result, the inability to work and move. Early detection of Parkinson's disease can help to cure it in time. A test that involves drawing a spiral on a sheet of paper could be used to detect Parkinson's disease in its early stages. Speed of writing and pen pressure while sketching are lower among Parkinson's patients, particularly those with a severe form of the disease.

Objective

The goal of this project is to develop an image classification algorithm to detect Parkinson's disease using images of spirals/waves/handwriting obtained during clinical exams and compare the results with other techniques that involve analysing the writing speed and pen pressure.

Note:

This project will involve research about the existing techniques on your part. In addition to this, you'll have to learn ML/DL concepts along the way to apply them for image classification.

To read

• https://www.upgrad.com/blog/using-convolutional-neural-network-for-image-classification/
• https://www.geeksforgeeks.org/image-classifier-using-cnn/
• https://machinelearningmastery.com/how-to-use-transfer-learning-when-developing-convolutional-neural-network-models/