IBM wants to detect Alzheimer earlier on the basis of machine learning

IBM has deployed machine learning models and artificial intelligence (AI) in the hope of detecting Alzheimer’s at an earlier stage. By diagnosing the disease earlier, people have more time to prepare themselves and their families. Patients are also eligible for medical trials.

In order to detect Alzheimer’s disease early on, a specific biological marker must be found in the spinal fluid. The concentration of the peptide amyloid-beta in that fluid changes decades before the first symptoms of the disease appear. People who already have mild cognitive impairments and an abnormal concentration of the peptide in their spinal fluid would be 2.5 times more likely to develop the disease.

However, the removal of spinal fluid is expensive and extremely invasive. However, IBM believes it has found a way to detect the disease at an early stage, without the need for invasive testing. It uses machine learning models developed by the Australian IBM Research team.

Operation

The models work on the basis of a set of algorithms, which identify a set of proteins in the blood. These proteins, in turn, can use machine learning to predict the concentration of amyloid-beta in the spinal fluid. The models could ultimately help predict the risk of Alzheimer’s disease with an accuracy of up to 77 percent.

In their own words, the new tests should be able to close the gap between early detection of the disease and clinical trials. IBM’s test is still at an early stage and is far from ready for use in the field of cognitive diseases. However, the team states that the new tests can eventually be used to detect other biological markers. These are not only markers of Alzheimer’s disease, but also of other diseases.