A simple drawing and hand movement test may help detect Parkinson's disease with an accuracy approaching 99%, according to a study by researchers in India.
Parkinson's is a neurological disorder that causes progressive damage to nerve cells, leading to symptoms such as tremors, slowness of movement, and muscle stiffness. The symptoms may worsen over time and affect the ability of those affected to live independently.
The diagnosis of the disease currently relies on assessing symptoms and neurological and physical examinations, and there is no single test that can definitively confirm the infection.
The researchers analyzed data from a previous study involving 66 people, including 31 with Parkinson's disease, who were asked to complete two drawing tasks. The tests involved drawing spirals, zigzags, and sharp-angled connected lines using a biometric pen that recorded hand movements during drawing.
The data showed that people with Parkinson's disease had more difficulty drawing straight lines than those without the disease. According to the researchers, this is related to the effect of tremors and motor control disorders on a patient's ability to steady a pen and maintain a consistent line.
The researchers used drawing and hand movement data to train different artificial intelligence models, focusing on indicators such as line regularity, movement speed, pen pressure, and hand movement coordination.
Next, the team processed the results using an algorithm known as "SNAKE", with the goal of re-analyzing the drawings and identifying participants with Parkinson's disease.
According to the results, the algorithm was able to detect the disease with an accuracy of 98.95% when analyzing the zigzag line drawings, while the accuracy was 97.7% when analyzing the spatial patterns of the drawings.
Researchers believe that drawing and handwriting analysis may one day provide a simple and less invasive way to help detect Parkinson's disease, but the results do not mean that the test is a substitute for medical diagnosis.
It remains unclear how effective this method is at detecting the disease in its early stages, and the results need further verification through larger studies, especially since the current study included only 66 participants and the algorithm was not tested on an independent group of participants.
The study was published in the journal Discover Computing.
what is your opinion on this news ?do let us know in the comments on tha comment box
if you liked this news article please share on friends & family & don't forget to follow on website & social media
