AI used to predict the most common type of dementia
Alzheimer’s disease is the most common type of dementia.
What you need to know:
- Some of the risk factors that the researchers looked out for during their analysis are high blood pressure, dizziness, abnormal stool contents, but more factors varied depending on one’s sex.
- For women for instance, osteoporosis (a condition that makes the bone weak) was a high risk link for them to develop Alzheimer’s disease in future.
- Men who had either enlarged prostate or a history of erectile dysfunction were highly susceptible to the disease.
Scientists have found a way to predict Alzheimer’s disease seven years prior to its onset using artificial intelligence (AI).
The team from the University of California in the US used machine learning and assessed more than five million medical records between 1980 and 2021.
The study published in the journal Nature Aging shows that the model application includes determining early disease risk in primary care settings before at-risk patients go for the time-consuming and costly diagnoses.
“This can aid in identification of at-risk patients for follow-up or inclusion in early intervention or trials, with the one-day prior model as suggesting possible Alzheimer’s disease onset to be considered at that visit to facilitate earlier diagnosis,” say the researchers.
“These findings potentially support hypotheses suggesting Alzheimer’s disease can be associated with general aging or frailty.”
To identify early clinical predictors that may be biologically relevant for Alzheimer’s disease diagnosis, the scientists trained machine learning models on individuals by matching pre-identified links such as demographics and hospital visits related to some of the risk factors that are known.
“This is a great example of how we can leverage patient data with machine learning to predict which patients are more likely to develop Alzheimer’s, and also to understand the reasons why that is so,” explains Marina Sirota, a computational health scientist at the University of California.
Some of the risk factors that the researchers looked out for during their analysis are high blood pressure, dizziness, abnormal stool contents, but more factors varied depending on one’s sex.
For women for instance, osteoporosis (a condition that makes the bone weak) was a high risk link for them to develop Alzheimer’s disease in future.
Men who had either enlarged prostate or a history of erectile dysfunction were highly susceptible to the disease.
Alice Tang, a bioengineer who was part of the study, said their model is a first step towards using AI on routine clinical data.
“It will be used to not only to identify risk as early as possible, but also to understand the biology behind it,” says Tang
“It is the combination of diseases that allows our model to predict Alzheimer’s onset. Our finding that osteoporosis is one predictive factor for females highlights the biological interplay between bone health and dementia risk,” she explains.
Besides predicting the risk of getting the disease, the scientists and researchers also explored the science behind some of the identified links like Osteoporosis for the women.
They found out that a variant in the gene known as MS4A6A was linked to it and this will help in further research on the bone disorder.
“These findings of osteoporosis as a potential sex-specific predictor of AD, with shared relationships through MS4A6A, is an unknown and unexpected result identified from single-hypothesis-driven follow-up from our prediction models,” shows the study
The scientists say that should the AI model be adopted for use in hospitals, there needs to be regular updates based on different factors.
“Due to changing patient demographics and societal factors, prediction models should be continuously trained, updated and evaluated if implemented in the clinical setting to ensure effective utilisation and account for biases that may have been learned from the data,” they explain.
