Tremor Detect

Abstract

The era of information technology brings with it a plethora of solutions in all fields especially if we talk about medical. Artificial Intelligence has brought some revolutionary changes in the medical field leaving the world astonished with its results. Parkinson’s Disease (PD) is a progressive, chronic illness, with the cause unknown and no cure. Treatment is effective only if the disease is caught early enough and early tremors typically begin in the fingers and hands. Different tests have been designed to detect tremors in hand. Two of the most common tests are the Spiral Test and the Wave Test. The traditional approach requires patients to draw either a spiral or a wave pattern on a paper. The presence of tremors is indicated by the irregular and shaky patterns drawn by the patient. The problem with this approach is that it is qualitative in nature, non - shareable and subjective to the doctor’s opinion. The idea is to develop a responsive web application that uses machine learning and digital image processing techniques. The application will provide digitized versions of the Spiral Test and the Wave Test and will predict whether a user has Parkinson’s Disease or not based on the digital input provided by the user. The digital inputs will be processed and quantified, and Parkinson’s prediction will be done through trained and tested machine learning models. These digitized tests enable users to take test from home. This also eliminates the expenses associated with routine clinical visits. Furthermore, these tests take a quantitative and standard approach thus removing the subjective aspect of the paper tests.