AI Developed Blood Tests May Be Able To Predict Parkinson’s Disease

 by Natalie M

According to research being done at UCL and University Medical Center Goettingen, a biomarker blood test has been developed using AI that can help diagnose and predict the likelihood of Parkinson's disease.

A progressive disorder which has symptoms consisting of tremors, slowness of movement and memory problems, Parkinson’s is characterised by a build of the alpha-synuclein protein. This is caused by the death of dopamine-releasing neurons in the substantia nigra in the brain, which controls movement. There are no current treatments for the disease but patients are currently being given dopamine replacement therapy; but researchers believe that early prediction and diagnosis of the disease would be highly beneficial to helping develop treatments that could stop/slow Parkinson's by protecting the neurons being affected.

Published in Nature Communications, the research described using machine learning (as a branch of AI) combined with mass spectrometry-based proteomics to analyse a panel of 8 blood-based biomarkers in both controls and individuals with Parkinson’s. It found that the concentrations of the 8 proteins are altered in patients with Parkinson’s. Thus, they used the panel and were able to use these biomarkers as a diagnostic test with 100% accuracy. 

To further investigate predicting the likelihood of a person developing Parkinson’s Disease, they analysed the blood from 72 patients with Rapid Eye Movement Behaviour Disorder (iRBD) - a sleep disorder presenting as patients physically enacting their vivid and often unpleasant dreams subconsciously. 75-80% of these people go on to develop synucleinopathies (neurodegenerative disorders that are characterised by the abnormal accumulation of alpha-synuclein) - which include Parkinson’s. The AI tool identified that 79% of the iRBD patients had the same profile as someone with Parkinson’s. The patients were later followed up over a 10-year period and so far the AI predictions have matched the clinical conversion rate, correctly predicting 16 patients as going on to develop Parkinson’s and also doing this 7 years before being symptomatic. The researchers are continuing to follow up on those who were predicted to have Parkinson’s to verify the accuracy of the test further.

This research is highly beneficial as it can be used as a non-invasive diagnostic tool for Parkinson’s in the future (which is in demand as imaging is no yet good enough to see what exactly is yapping in brain cells to definitively diagnose and samples of the brain cannot be taken as easily do at all - unlike other organs) and finding biomarkers that can be identified and measured is non-invasive (compared to lumbar puncture which is used more in clinical trials of treatments for Parkinson’s). According to Professor David Dexter (the Director of Research at Parkinson’s UK), "With more work, it may be possible that this blood based test could distinguish between Parkinson’s and other conditions that have some early similarities, such as Multiple Systems Atrophy or Dementia with Lewy Bodies. This is an important next step… The findings add to an exciting flurry of recent activity towards finding a simple way to test for and measure Parkinson’s." However, there are other diagnostic methods being tested including skin swabbing, eye scans and the lumbar puncture test so blood testing is not the only method being trialled for Parkinson’s.