IBM’s new research leverages AI to predict Alzheimer’s
IBM’s new research leverages AI to predict Alzheimer’s

A team at IBM Research Australia is developing a blood test to detect early stage Alzheimer’s disease.

Led by researcher Ben Goudey, the team used artificial intelligence (AI) technology to identify a set of proteins in blood that could predict amyloid-beta concentration in spinal fluid.

Previous studies indicated that amyloid-beta is a biological marker associated with Alzheimer’s.

It was found that people with mild cognitive impairment that were predicted to possess an abnormal amyloid concentration in their spinal fluid were 2.5 times more likely to develop the disease.

The new test identifies this marker to predict the risk of patients developing the neurodegenerative condition with up to 77% accuracy. The test also has applications in selecting patients for drug trials.

Data from the research has been published in Scientific Reports.

“As our population lives longer, neurodegenerative diseases such as Parkinson’s, Alzheimer’s and Huntington’s are affecting millions of people around the world.”

Ben Goudey’s team also noted that the machine learning approach can be applied to model other spinal fluid-based biomarkers. IBM researchers intend to present their work on a separate blood test for another key Alzheimer’s biomarker called tau.

In the company’s blog post, Goudey said: “As our population lives longer, neurodegenerative diseases such as Parkinson’s, Alzheimer’s and Huntington’s are affecting millions of people around the world.

“At IBM Research, our mission is to use AI and technology to understand how to help clinicians better detect and ultimately prevent these diseases in their early stages.”

Commonly, Alzheimer’s is diagnosed upon observing significant memory loss. Despite numerous clinical trials, the condition currently lacks a cure or disease-modifying therapies.

Failure rates in trials are believed to be due to the recruitment of patients that are in late stages of the disease when their brains have already suffered significant tissue loss that cannot be easily repaired. Therefore, multiple studies are being conducted to detect the disease in earlier stages.

Recently, a Duke University team studied the use of a non-invasive technology called optical coherence tomography angiography (OCTA) for the detection of Alzheimer’s through eye exams.

The new approach is based on the team’s finding that blood vessels in the retina are different in people with Alzheimer’s. An OCTA scan can reveal changes in capillaries of the retina even before blood vessel changes show up on a brain scan.

In January this year, a team of US and German researchers unveiled a blood test aimed to identify the disease before symptoms appear, while researchers at Brigham and Women’s Hospital in the US announced a similar test in December last year.

Related: IBM Watson Health announces slew of AI-based research partnership deals

RELATED STORIES