Original title: Artificial intelligence applications in Parkinson's disease via retinal imaging
Ali Jafarizadeh, Hamidreza Ashayeri, Hadi Vahedi, Parsa Khalafi +5 · March 2026
Parkinson's disease is expected to rise as the population ages. Catching it early matters because timely detection might slow it down and lower long-term complications. Researchers reviewed 35 years of evidence to see if artificial intelligence can spot early signs of the disease in the tiny blood vessels at the back of the eye, known as the retina. Changes in these retinal vessels act as physical signs of brain degeneration. The researchers found that AI computer programs can scan images of these eye vessels to find patterns tied to Parkinson's. Across nineteen analyzed studies, AI systems succeeded at three main jobs: identifying whether someone has the disease, mapping out retinal blood vessels, and predicting a patient's risk. Overall, combining AI with eye scans shows strong potential for spotting the disease earlier and more accurately than regular clinical checks alone.
You are trying to tell if a house has termites by looking at the paint on the walls. The paint might show faint bubbles that are easy to miss. A computer trained on many houses could learn to spot those bubbles. Here, the AI learns to spot early Parkinson's from the pattern of retinal blood vessels, which are like the paint on the walls of the brain.
Parkinson's disease is hard to catch early, but this research shows that AI analysis of eye scans could flag the disease before obvious symptoms appear. That could give patients a head start on treatment and turn a routine eye exam into a screening tool for millions. Earlier detection may delay progression and reduce long-term complications.
This is a review of existing studies, not a new experiment. The results come from specific datasets that may not reflect real-world diversity. The AI models have not been tested in clinical practice. The review also notes that only 19 studies met criteria, so the evidence base is still small.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Artificial intelligence applications in Parkinson's disease via retinal imaging — Ali Jafarizadeh, Hamidreza Ashayeri, Hadi Vahedi et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.