A new AI tool may predict Parkinson's disease up to 15 years before diagnosis

An artificial intelligence (AI) model has been trained to detect Parkinson’s disease up to 15 years before diagnosis.

Parkinson’s, the fastest-growing neurological disorder across the world, is caused by the loss of nerve cells in the brain. Symptoms include body tremors, slow movement, body stiffness and balance issues.

Other symptoms including depression, problems sleeping, constipation and loss of smell can appear decades before the more typical physical and cognitive signs. There are currently no tests for the disease – diagnosis is based on symptoms, a physical examination and medical history.

However, a team from the University of New South Wales (UNSW) and Boston University trained their AI model to detect the disease in blood samples, using data generated by a study of 41,000 participants investigating cancer and nutrition.

The AI model was given blood data from a random selection of 39 participants who went on to develop Parkinson’s and the same number of control patients. By measuring metabolites in the blood, the AI was able to identify those who later developed Parkinson’s with 96% accuracy, and up to 15 years before a clinical diagnosis. 

Studies show diagnosis of Parkinson’s by current clinical methods ranges from 65% to 93%.

Writing in The Conversation, Diana Zhang and Assistant Professor William Alexander Donald said: ‘Parkinson’s is not genetic, has no specific test and cannot be accurately diagnosed before motor symptoms appear.

‘Overall, AI could detect Parkinson’s disease with up to 96% accuracy. The AI tool also helped us identify which chemicals or metabolites were likely linked to those who later developed the disease.’

Metabolites are produced by the body when it breaks down food, drugs and other substances. A difference in concentrations of particular metabolites can indicate disease – or resistance to it.

‘Our research identified a chemical, likely a triterpenoid, as a key metabolite that could prevent Parkinson’s disease,’ said the UNSW team. ‘The abundance of triterpenoid was lower in the blood of those who developed Parkinson’s compared to those who did not.

‘A synthetic chemical (a polyfluorinated alkyl substance) was also linked as something that might increase the risk of the disease. This chemical was found in higher abundances in those who later developed Parkinson’s.’

While the team highlighted the need for further research to validate the results, including on larger populations, they are hopeful it will improve patient quality of life by detecting the disease earlier – also potentially reducing health care costs. They have already made the tool publicly available.

The study is published in the journal ACS Central Science.

Source: Read Full Article