The study uses machine learning technology to study the structural features of the brain, including in regions not previously associated with Alzheimer’s disease. The advantage of the method lies in its simplicity and in the fact that it allows you to identify the disease at an early stage, when it can be very difficult to diagnose.
While there is no cure for Alzheimer’s disease, early diagnosis helps patients. This allows them to access help and support, receive treatment to manage their symptoms, and plan for the future. Being able to accurately identify patients at an early stage of the disease will also help researchers understand the changes in the brain that trigger the disease, and will facilitate the development and testing of new treatments.
The study was published in Nature Portfolio, Communications Medicine.
Alzheimer’s disease is the most common form of dementia, affecting over half a million people in the UK. Although most people develop Alzheimer’s after the age of 65, it can also develop in people younger than that age. The most common symptoms of dementia are memory loss and difficulty with thinking, problem solving, and language.
Physicians now use a range of tests to diagnose Alzheimer’s, including memory and cognitive ability tests and brain scans. The scan is used to check for protein deposits in the brain and to shrink the hippocampus, an area of the brain associated with memory. All of these tests can take several weeks to organize and process.
The new approach requires only one of them – magnetic resonance imaging (MRI) of the brain, performed on a standard 1.5 Tesla machine, which is usually available in most hospitals.
The researchers adapted an algorithm developed to classify cancerous tumors and applied it to the brain. They divided the brain into 115 regions and identified 660 different characteristics, such as size, shape, and texture, to evaluate each region. They then trained the algorithm to determine where changes in these characteristics could accurately predict the presence of Alzheimer’s disease.
Using data from the Alzheimer’s Neuroimaging Initiative, the team tested their approach on brain scans of more than 400 early and advanced Alzheimer’s patients, healthy individuals, and patients with other neurological conditions, including frontotemporal dementia and Parkinson’s disease. They also tested it against more than 80 patients undergoing diagnostic tests for Alzheimer’s at the Imperial College Healthcare NHS Trust.
They found that in 98 percent of cases, the MRI-based machine learning system could accurately predict whether or not a patient had Alzheimer’s disease. She was also able to distinguish between early and late stages of Alzheimer’s disease with reasonable accuracy in 79 percent of patients.
Professor Eric Aboagier of the Imperial Department of Surgery and Oncology, who led the study, said: “Currently, no other simple and widely available methods can predict Alzheimer’s disease with this degree of accuracy, so our study is an important step forward. Many patients who come to the clinics memory patients with Alzheimer’s disease have other neurological diseases, but even in this group, our system was able to distinguish between patients with Alzheimer’s disease and those without it.
“Waiting for a diagnosis can be a terrifying experience for patients and their families. If we could reduce waiting times, simplify the process of making a diagnosis, and reduce some of the uncertainty, that would help a lot.” Our new approach can also identify patients in the early stages of the disease for clinical trials of new drugs or lifestyle changes, which is currently very difficult to do.”
The new system found changes in areas of the brain not previously associated with Alzheimer’s disease, including the cerebellum (the part of the brain that coordinates and regulates physical activity) and the ventral diencephalon (associated with the senses, sight and hearing). This opens up new potential avenues to explore these areas and their association with Alzheimer’s disease.