PositiveMed

The Positive Side of Medicine

Predicting The Risk of Alzheimer’s Disease With AI

Predicting The Risk of Alzheimer’s Disease With AI

Share This Post

These days, armed with medical and scientific knowledge and with the newest digital tools, we are going to great lengths to fight every illness. No wonder that new Artificial Intelligence (AI) medical algorithms arise often. Alzheimer’s disease is the primary cause of dementia worldwide, with one in 10 people aged 65 and older exhibiting Alzheimer’s dementia. The newest AI systems can accurately predict and diagnose the risk of it. Let’s browse the scientists’ brand-new Alzheimer’s disease with AI, interesting discoveries. 

Predicting The Risk of Alzheimer’s Disease With AI

Current knowledge on treating Alzheimer’s disease

Pharmaceutical companies have invested hundreds of billions in Alzheimer’s research, however, the field has had little success over the years, with 146 unsuccessful attempts to develop drugs that prevent or treat the disease, between 1998 and 2017. Only four new medicines received approval and solely to treat symptoms. More than 90 drug candidates are still in development. Now, the Massachusetts Institute of Technology (MIT) researchers are looking to partner with pharmaceutical firms to get the model implemented in real-world clinical trials for Alzheimer’s. 

Alzheimer’s disease and AI

With the use of Artificial Intelligence, researchers from MIT took time to develop tools to determine whether patients at high risk of Alzheimer’s disease would experience cognitive decline. The algorithm was able to predict patient cognition test scores for up to two years in the future. The deep learning algorithm, developed by researchers at the Boston University School of Medicine, uses a combination of brain magnetic resonance imaging (MRI) testing to measure cognitive impairment, along with data on age and gender, which helps to accurately predict the risk of Alzheimer’s Disease. For new participants, researchers developed a second model that is personalized for each patient and continuously updates risk scores. 

Meta-learning scheme

To optimize the results of the personalized models, and to use newly recorded data at its best, researchers invented a “meta-learning” scheme. It learns to automatically choose which type of model works best, for any given participant, at any given time. While meta-learning has been used before for computer vision and machine translation tasks, the team said this is the first time when it’s used for Alzheimer’s. The meta-learning scheme helped to reduce the error rate for future predictions by fifty percent. “We wanted to learn how to learn with this meta-learning scheme. It’s like a model on top of a model that acts as a selector, trained using meta-knowledge to decide which model is better to deploy” – says Oggi Rudovic, a Media Lab researcher.

MRI scans at use. Expanding the use of neuroimaging data

How to create a useful database for treating Alzheimer’s medical predictions? The researchers collected MRI scans of the brain, demographics, and clinical information of individuals with Alzheimer’s disease as well as ones with normal cognition. Afterward, they developed a novel deep learning model to accurately predict Alzheimer’s disease status on the other independent cohorts.

This study has broad implications for expanding the use of neuroimaging data (among which MRI scans) to accurately detect risks. Researchers first trained a population model on an entire dataset that included clinically significant cognitive test scores and other biometric data from Alzheimer’s patients and healthy individuals between biannual doctor’s visits. The model analyzes data learning patterns that can help predict how patients will score on cognitive tests.

Preparing for the future

Why do we need these models? They could help with selecting candidate drugs and participants for clinical trials, and allow patients and their loved ones prepare for rapid cognitive decline. “Being able to accurately predict future cognitive changes can reduce the number of visits the participant has to make, which can be expensive and time-consuming. Apart from helping develop a useful drug, the goal is to help reduce the costs of clinical trials to make them more affordable and done on larger scales” – said Oggi Rudovic, a Media Lab researcher.

This algorithm can generate interpretable and intuitive visualizations of individual Alzheimer’s disease risk. These steps could be extended to other organs in the body, leading to improved drug development and care delivery. We can also already assume that it’s possible to diagnose other degenerative diseases with the help of AI. “If we have accurate tools to predict the risk of Alzheimer’s disease […], that is readily available and which can use routinely available data, such as a brain MRI scan, then they have the potential to assist clinical practice […]” – says Vijaya Kolachalama, Ph.D., assistant professor of medicine at Boston University School of Medicine.

More To Explore

Scroll to Top