Machine learning offers a nuanced view of the stages of Alzheimer’s — ScienceDaily

A Cornell-led collaboration used machine learning to identify the most accurate tools and timelines for predicting Alzheimer’s disease progression in people who are cognitively normal or have mild cognitive impairment.

Modeling showed that it was easier and more accurate to predict future decline in dementia in individuals with mild cognitive impairment than in cognitively normal or asymptomatic individuals. At the same time, the researchers found that the estimates for cognitively normal subjects were less accurate for longer time horizons, but the reverse was true for those with mild cognitive impairment.

The modeling also showed that magnetic resonance imaging (MRI) is a useful prognostic tool for people in both stages, while tools that monitor molecular biomarkers such as positron emission tomography (PET) scans are more useful for people with mild cognitive impairment.

The team’s paper entitled “Machine-Learning-Based Multimodal Prediction of Future Decline in Alzheimer’s Disease: An Empirical Study,” published Nov. 16 in PLOS ONE. The lead author is Batuhan Karaman, a PhD student in electrical and computer engineering.

Alzheimer’s disease can take years, sometimes decades, to progress before a person shows symptoms. Once diagnosed, some people regress rapidly, while others may live with mild symptoms for years, making the rate at which the disease progresses difficult to predict.

Senior author Mert Sabuncu, associate professor of electrical and computer engineering in the School of Engineering, said, “When we can confidently say that someone has dementia and electrical engineering in radiology at Weill Cornell Medicine, it is too late.

“We need to be able to catch Alzheimer’s disease really early and be able to tell who will progress faster and who will progress slower, so we can categorize and activate different risk groups,” Sabuncu said. whatever treatment options we have.”

Clinicians often focus on a single “time horizon” (usually three or five years) to predict the progression of Alzheimer’s in a patient. According to Sabuncu, whose lab specializes in the analysis of biomedical data, the time frame may seem arbitrary—imaging data with a particular emphasis on neuroscience and neurology.

Sabuncu and Karaman partnered with longtime collaborator and co-author Elizabeth Mormino of Stanford University to use neural network machine learning that can analyze five years of data on individuals who are cognitively normal or have mild cognitive impairment. The data collected in a study by the Alzheimer’s Disease Neuroimaging Initiative covered everything from an individual’s genetic history to PET and MRI scans.

“What we’re really interested in is can we look at this data and tell whether a person will progress in the coming years?” said the Soapmaker. “And more importantly, can we do a better job of predicting when we combine all the tracking data points we have on individual subjects?”

Researchers have discovered several notable patterns. For example, it is much easier to predict that a person will go from asymptomatic to exhibiting mild symptoms for a time horizon of one year compared to five years. However, predicting whether a person will decline from mild cognitive impairment to Alzheimer’s dementia is most accurate on a longer timeline, the “sweet spot” being about four years.

“This may tell us something about the underlying disease mechanism and how it develops transiently, but that’s something we haven’t studied yet,” Sabuncu said. Said.

Modeling showed that, regarding the effectiveness of different types of data, MRI scans are most informative for asymptomatic cases and are particularly helpful in predicting whether someone will develop symptoms in the next three years, but less helpful in predicting for people with mild cognitive impairment. . PET scans that measure certain molecular markers such as amyloid and tau proteins appear to be more effective when a patient develops mild cognitive impairment.

An advantage of the machine learning approach is that neural networks are flexible enough to work despite missing data, such as patients who may have skipped an MRI or PET scan.

In his future work, Sabuncu plans to further modify the modeling so that it can process full imaging or genomic data rather than summary measurements to gather more information that will improve predictive accuracy.

The research was supported by the National Institutes of Health’s National Library of Medicine and the National Institute on Aging and the National Science Foundation.

Story Source:

materials provided by Cornell University. Original written by David Nutt, courtesy of the Cornell Chronicle. Note: Content can be edited for style and length.

Leave a Reply

Your email address will not be published. Required fields are marked *