New technologies make it possible to find out if there is a risk of contracting the disease even if there are no pre-symptoms.
Alzheimer’s disease can last for years, sometimes decades, before symptoms appear. Once the disease is diagnosed, some individuals degenerate rapidly, but others can live with mild symptoms for years, making it difficult to predict how fast the disease will progress.
A collaboration led by Cornell University in New York has used machine learning to identify the most accurate means and timing of anticipating the progression of Alzheimer’s disease in cognitively normal people or people with mild cognitive impairment.
Modeling has shown that predicting future decline to dementia for individuals with mild cognitive impairment is easier and more accurate than for cognitively normal or asymptomatic individuals. At the same time, the researchers found that the predictions were highly accurate for subjects with mild cognitive impairment.
Modeling has also shown that MRI is a useful prognostic tool for people in both stages (cognitively normal individuals, individuals with mild cognitive impairment) while tools that track molecular biomarkers, such as positron emission tomography ( PET ), are most useful for people with mild cognitive impairment.
“By the time we can say for sure that someone has dementia, it’s too late. The brain has already suffered a lot of damage, and it is irreversible damage. We need to understand Alzheimer’s disease early and understand who will progress quickly and who will progress slowly, so that we can stratify the different risk groups and use the available treatment options.
Advertisement
said senior author Mert Sabuncu, an associate professor of electrical and computer engineering at the Weill Cornell Medicine College of Engineering. “Regarding the effectiveness of different types of data, modeling has shown that MRI is most informative for asymptomatic cases and is particularly useful for predicting whether a person will develop symptoms in the next three years, but is less useful for predicting people with mild cognitive impairment. Once the patient has developed mild cognitive impairment, PET scans, which measure certain molecular markers such as amyloid and tau proteins, appear to be more effective.”
- Machine learning based multi-modal prediction of future decline toward Alzheimer’s disease: An empirical study. (journals.plos.org)
