Mount Sinai Health System’s recently launched digital pathology offshoot, PreciseDx, was created to support prostate and breast cancer diagnoses. Now, early detection of Parkinson’s disease has also been found to be possible through its advanced AI programs – even before the presentation of severe symptoms.
The AI of PreciseDx was initially built to evaluate the form and structure of cells on slides populated with biopsied tissue in order to trace early-stage indicators of cancer, helping to gauge an individual’s risk of tumor development or recurrence after treatment. When used to analyze nerve samples from salivary glands, Mount Sinai researchers discovered the ability to seek out protein clumps called Lewy bodies, which are associated with Parkinson’s.
"Traditionally, pathology grading systems look at a few morphology components to make a diagnosis,” said Dr. John Crary, a Professor in the Pathology, Neuroscience, and AI departments of Mount Sinai’s Icahn School of Medicine. “Unlike any human-powered grading method, PreciseDx's AI Morphology Feature Array can examine thousands of different features and leverage those relationships between them." A study performed in conjunction with The Michael J. Fox Foundation for Parkinson's Research found that the AI could reach high levels of accuracy. Outperforming human pathologists, the technology can provision false-positive and false-negative outcomes at a 1% rate.