An AI Scans Eyes to Spot Parkinson’s Years Before Diagnosis – DNyuz

An AI Scans Eyes to Spot Parkinson’s Years Before Diagnosis

When it comes to neurodegenerative diseases like Alzheimer’s or Parkinson’s, the earlier you can treat it, the better. This can not only give patients the chance to delay disease by making lifestyle changes, but also provide clinicians with valuable information that will help them guide their treatment.

These are hard diseases to catch early, though. Most diagnoses happen when the disease has made itself known through clear symptoms like body tremors or speech changes. It’s then a race against time to treat symptoms, and prevent them from worsening.

However, that might soon all change with some new research published Monday in the journal Neurology that found that signs of Parkinson’s disease can be spotted years before diagnosis and before symptoms begin to show by scanning patients’ eyeballs. The technique utilizes an AI model to detect the presence of the neurodegenerative disease in patients on average seven years before clinical diagnosis.

“Finding signs of a number of diseases before symptoms emerge means that, in the future, people could have the time to make lifestyle changes to prevent some conditions arising, and clinicians could delay the onset and impact of life changing neurodegenerative disorders,” lead author Siegfried Wagner, a researcher at the University College London Institute of Ophthalmology, said in a press release.

Scientists have long known that eye scans can reveal a lot about human health including the likelihood of developing cardiovascular diseases, diabetes, high blood pressure, and more. It turns out that neurodegenerative disorders can also be detected through eye scans.

Specifically, the UCL team built off of previous research that they had done on the ganglion cell-inner plexiform layer (GCIPL), a membrane in the retina. The researchers found that GCIPL loss was associated with an increased risk for neurodegenerative diseases like Parkinson’s.

The scientists combined these insights into a machine-learning model trained using AlzEye – a large dataset of retinal images collected in order to diagnose dementia. In order to reproduce their results, the AI was trained on data from U.K. Biobank which contained diagnostic information about healthy volunteers.

Using the datasets, the AI was able to identify very subtle signs of GCIPL atrophy in eye scans and, therefore, their increased likelihood of developing Parkinson’s. These markers were detected up to seven years before diagnosis of the disease.

“This work demonstrates the potential for eye data, harnessed by the technology to pick up signs and changes too subtle for humans to see,” Alistair Denniston, a consultant ophthalmologist at University Hospitals Birmingham who wasn’t involved in the study, said in a press release. “We can now detect very early signs of Parkinson’s, opening up new possibilities for treatment.”

More research is needed to understand why exactly GCIPL becomes thinner as a patient develops Parkinson’s. The new AI-based detection method is an important tool to fight a devastating and pernicious disease.

The post An AI Scans Eyes to Spot Parkinson’s Years Before Diagnosis appeared first on The Daily Beast.