IBM Sets hope for Alzheimer’s Disease Diagnoses with Machine Learning

by March 12, 2019

Machine Learning

Alzheimer’s disease starts slowly worsening over time, is the cause of 60–70% of cases of dementia associated with a decline in memory and other thinking skills severe enough to reduce a person’s ability to perform everyday activities.

At the onset, the most common early symptom of Alzheimer’s is the difficulty in remembering recent events. The diagnosis is as difficult as the ailment. In a recent development, IBM has claimed that it has developed a machine learning model that can be programmed to aid clinicians to predict the severity of the Alzheimer’s disease additionally in showing its progression. The company’s Australian arm, IBM Research – Australia, undertook a research publishing its findings in the journal titled Scientific Reports.


The Research and its Findings

IBM Australia came up with the details explaining how machine learning can be deployed to diagnose Alzheimer’s helping patients and family members to predict the severity.

The research was conducted by Ben Goudey, a researcher at the Genomics Research Team of IBM Research and the Research team at IBM Australia. The study found that machine learning has been deployed to identify a set of proteins in the blood that has the ability to successfully predict the concentration of the biomarker amyloid-beta (amyloid-β) in spinal fluid. The ground-breaking research can help predict the risk of Alzheimer’s disease long before it gets detected is a boon to patients and their families.

Research suggests a peptide called amyloid-beta changes long before memory loss occurs. Analysing the concentration of this peptide from an individual’s spinal fluid could highlight the risk decades of Alzheimer’s long before in advance.

Accessing spinal fluid is an invasive and expensive procedure as, Goudey wrote in his post “Hence, there is a strong effort in the research community to develop a less invasive test, such as a blood test, that can yield information about Alzheimer’s disease risk.”

The research paper, dated March 11, 2019, collated by Goudey and his team has been published in the Scientific Reports. The research is still in its initial stages, might help clinicians to predict the risks with an accuracy of up to 77 percent improving the selection of people for drug trials. As by Goudey explains,

“Individuals with mild cognitive impairment who were predicted to have an abnormal concentration of amyloid in their spinal fluid were found to be 2.5 times more likely to develop Alzheimer’s disease”

Goudey’s latest research has potentially identified proteins in the bloodstream that may help indicating the concentration of amyloid-β in the spinal fluid. Thus, less invasive tests such as a blood test can be conducted instead of people having to go through the more invasive spinal fluid test to detect the ailment.

Moreover, Goudey and his team have also been working on a blood test to analyse another important biomarker of Alzheimer’s disease, known as tau. The studies will be presented at the end of this month at the 14th International Conference on Alzheimer’s and Parkinson’s Diseases in Lisbon.


The Brutal Alzheimer’s

Alzheimer’s is a brutal disease not just for sufferers, but their families as well. The disease currently has no cure, leading to a loss of memory, confusion, and difficulty completing tasks which were once familiar to the person.

Early diagnosis of Alzheimer’s helps the patient and their family to prepare the sufferer and their loved ones as much as possible before degeneration becomes massive. Official diagnosis also helps the patient available for medical trials with the hope to find a full cure someday.

This research, when it comes to reality can be a boom in the field of medicine take one step forward to the potential of detecting Alzheimer’s disease at a budding stage through machine learning.