Parkinson’s Disease (PD) is an illness of the central nervous system that affects movement. Dopamine levels in the brain drop below safe levels resulting in tremors, slow movement, stiffness, and loss of balance.
PD cannot be cured at this time. Some treatments can help, especially when caught early. The problem is that catching PD early is difficult. It requires a medical diagnosis involving laboratory testing and imaging. There are no biomarkers for early detection. PD can last for years.
https://press.rsna.org/timssnet/media/pressreleases/14_pr_target.cfm?ID=2229 The Radiological Society of North America published a report (Poster IN-1A-07) last week identifying the potential use of eye exams to diagnose PD. Artificial Intelligence (AI) is using a process called vector machine learning to detect similarities and suggest outcomes.
The blood vessels in the retina transition into a distinct pattern when dopamine levels in the brain decline. AI can assess and evaluate the difference between healthy eyes and eyes that are shifting into a non-healthy condition.
https://www.medscape.com/viewarticle/941700?src=rss The overall process is simple and quick. Detection and diagnosis are affordable and portable. The J. Crayton Pruitt Department of Biomedical Engineering’s Smart Medical Informatics Learning and Evaluation Lab (SMILE) collected fundus eye images of healthy and PD patients.
The fundus is the inside, back portion of the eye. The retina, macula, blood vessels, and more make up the eye fundus. A special camera is used that photographs the inside of the eye through the pupil.
80% of the images were used to create the machine language used in AI. The other 20% were used as the baseline to test the program algorithms.
https://www.ajmc.com/view/how-an-eye-exam-may-improve-early-diagnosis-of-parkinson-disease 85% accuracy was achieved in identifying healthy eyes and eyes from PD patients. The test can determine if someone has PD but cannot tell how long the disease has been present. The downside is that the test is effective when 80% of the dopamine cells in the brain have decayed.
This AI approach to diagnosing PD is being evaluated to determine the effectiveness of different drug treatments used to manage PD and other eye diseases.
This is a great step forward to confirm a serious health issue early enough that drug treatments might provide a better quality of life. I can see this approach being used as a diagnostic tool for other eye diseases.
When thousands of images have been added to the database, it is possible that better predictions can be made earlier. It might even be able to identify diseases not associated with the eyes, such as Alzheimer’s Disease.
Live Longer & Enjoy Life! – Red O’Laughlin – RedOLaughin.com