AI analyzes gray matter loss to predict onset of Alzheimer’s disease

Advances in artificial intelligence promise to open up all kinds of possibilities when it comes to healthcare, analyzing medical images for signs of problems is already proving to be a great strength of technology. Scientists at the University of Cambridge have shown that a new type of machine learning algorithm can detect structural changes in the brain indicating early dementia and can be combined with standard memory tests to calculate the likelihood that a person develops Alzheimer’s disease.

While clinicians are good at analyzing PETs, MRIs, or other types of medical images for irregularities that may be related to disease, artificial intelligence promises to overload these forms of diagnosis. The power of modern computing makes machine learning algorithms very well suited to detecting subtle changes in, for example, brain tissue, which would escape even the highly skilled eyes of today’s physicians.

We have seen this very promising technology in detect ventricular dysfunction, a major precursor to heart failure, for example, or more effectively detect malignant tissue in lung nodules which could be a sign of cancer. Along with these advances, we’ve also seen scientists make progress in detecting the onset of Alzheimer’s disease, potentially long before symptoms appear.

A compelling example came in 2018, when scientists at University College London demonstrated a new machine learning algorithm that could detect subtle patterns in dense brain imaging data that represent changes in glucose uptake. Tested on a small set of brain scans, the algorithm was then able to predict each case that would develop Alzheimer’s disease on average six years before their diagnosis.

Scientists at the University of Cambridge pursued a similar possibility, but through a different physiological mechanism. The team trained their machine-learning algorithm on brain scans of patients who developed Alzheimer’s disease, through which they learned to detect structural changes related to the density of gray matter in the brain that denoted the stages of disease formation.

“We have trained machine learning algorithms to detect early signs of dementia simply by looking for patterns of gray matter loss – essentially, wear and tear – in the brain,” says Professor Zoe Kourtzi. “When we combine this with standard memory tests, we can predict whether an individual will exhibit a slower or faster decline in their cognition. We were even able to identify some patients who were not yet showing symptoms, but who developed Alzheimer’s disease. “

The algorithm was applied to brain scans of patients already with mild cognitive impairment and suffering from memory loss or impaired perception of language, vision or space. Combined with memory tests, the algorithm was found to be able to predict who would develop Alzheimer’s disease with 80% accuracy, and was also able to predict the rate of their cognitive decline.

“Ultimately, we hope to be able to identify patients as early as five to ten years before they show symptoms during a check-up. Kourtzi said.

Like the other examples mentioned above, this research is still in its early stages and it will take a lot more work to use it in the clinic, starting with the replication of the results in a cohort larger than the 80 patients. involved in this initial research project. Scientists are currently developing a clinical study for this, and although the algorithm is optimized for Alzheimer’s disease, they also hope to adapt it to detect other forms of dementia by identifying other unique patterns of structural change. in the brain.

“We have shown that this approach works in a research setting – now we need to test it in a ‘real world’ setting,” says study author Dr Timothy Rittman.

The team’s research paper, which has yet to be peer-reviewed, is available here.

Source: Cambridge University

Leave a Comment