Undoubtedly, diagnosing Parkinson’s disease is a challenge. Some procedures to identify signs of the disease have high costs, such as computed tomography and magnetic resonance imaging. In addition, millions of people worldwide do not have access to these technologies. According to the World Health Organization (WHO), currently, about 6.3 million suffer from the problem.
Since the first symptoms only manifest with the progression of the disease, early detection is another challenge. A new research from the United States has developed a cheaper way to diagnose the disorder: artificial intelligence applied to the retinal exam. The results were presented at the last annual meeting of the Radiological Society of North America (RSNA).
Learn about the study, how it was done and how the results achieved help to democratize access to healthcare.
Researchers at the University of Florida (USA) used the machine learning principle to create an artificial intelligence tool that learns to detect signs of Parkinson’s disease in retinal examinations. They trained the “support vector machine” with eye fundus images of patients with the disease and control participants without the disorder.
As the problem deteriorates nerve cells and, consequently, thins the retinal walls, a simple fundus examination can already diagnose it. The disease also damages the retinal microvasculature.
The results show that Parkinson’s disease can be diagnosed from changes in the retina. Currently, several studies prove that damage to the brain can be observed through the eyes .
“The most important finding was that a brain disease was diagnosed with a simple eye image. The diagnosis can be made in less than a minute and the equipment costs much lower than a computed tomography or magnetic resonance imaging ”, says Maximillian Diaz, the researcher in charge.
Undoubtedly, by applying machine learning techniques to the artificial intelligence system used in the retinal scan, scientists can diagnose Parkinson’s disease faster, more assertively, cheaply and accessibly.
Early detection, even before the first symptoms, allows the treatment to start as soon as possible and provide the patient more life-quality.
The results of the research can also help in a better understanding of the disease in the search for a cure and in ways to slow the evolution.
In addition, the researchers say the new artificial intelligence tool can be used to identify other diseases that damage the brain, such as Alzheimer’s disease and multiple sclerosis.
Follow the main researches related to the eyes on Phelcom’s blog.