Mount Sinai researchers have used new artificial intelligence methods to examine the structural and cellular features of human brain tissue to help identify the causes of Alzheimer’s disease and other related disorders. The research team found that studying the causes of cognitive impairment using an unbiased AI-based method – in contrast to traditional markers such as amyloid plaques – revealed unexpected microscopic anomalies that can predict the presence of cognitive impairment. These results were published in the journal Acta Neuropathologica Communications On September 20.
AI represents an entirely new paradigm for the study of dementia and will have a transformative impact on research into complex brain diseases, particularly Alzheimer’s disease. A deep learning approach has been applied to predict cognitive impairment, which is a challenging problem for which there is currently no human-made histological diagnostic tool.”
John Crary, MD, PhD, co-author, professor of pathology, molecular and cellular medicine, neuroscience, artificial intelligence and human health, Icahn School of Medicine at Mount Sinai
Mount Sinai’s team identified and analyzed the basic structure and cellular features of two brain regions, the medial temporal lobe and frontal cortex. In an effort to improve the level of postmortem brain assessment to identify signs of disease, the researchers used a poorly supervised deep learning algorithm to examine slice images of human brain anatomy tissue from a pool of more than 700 elderly donors to predict presence or absence. of cognitive impairment. A weakly supervised deep learning approach is able to handle noisy, limited, or inaccurate sources to provide signals for classifying large amounts of training data in a supervised learning environment. This deep learning model was used to quantify a decrease in Luxol’s fast blue staining, which is used to quantify myelin, the protective layer around the brain’s nerves. Machine learning models identified a signal of cognitive impairment associated with decreased amounts of myelin staining; scattered in an irregular pattern across the tissues; It is concentrated in the white matter that affects learning and brain function. Both sets of models the researchers trained and used were able to predict the presence of cognitive impairment with better accuracy than random guesses.
In their analysis, the researchers believe that the decreased intensity of coloration in specific brain regions identified by AI may serve as a scalable platform for assessing the presence of brain impairment in other associated diseases. The methodology lays the foundation for future studies, which could include dissemination of artificial intelligence models on a larger scale as well as further slicing of the algorithms to increase their predictive accuracy and reliability. Ultimately, the team said, the goal of this neurological disease research program is to develop better tools for diagnosing and treating people with Alzheimer’s disease and related disorders.
“Leveraging AI allows us to look at more disease-relevant features, which is a powerful approach when applied to a complex system like the brain,” said co-author Kurt W. Farrell, Ph.D., assistant professor of pathology, molecules and cell. human”. Existing medicine, neuroscience, artificial intelligence and human health, in Icahn Mount Sinai. “It is critical that further interpretation research be conducted in the areas of neuropathology and artificial intelligence, so that advances in deep learning can be translated to improve diagnostic and treatment approaches for Alzheimer’s disease and related disorders in a safe and effective manner.”
Lead author Andrew Mackenzie, MD, PhD, and associate resident chair of research in the Icahn Division of Psychiatry at Mount Sinai added: “Interpretation analysis was able to identify some, but not all, of the signals that AI models used to make predictions about cognitive impairment. As a result, additional challenges remain to deploy and interpret these powerful deep learning models in the field of neuropathology.”
Also contributing to this research were researchers from the University of Texas Health Science Center in San Antonio, Texas, University of Newcastle in Tyne, UK, Boston University School of Medicine in Boston, and UT Southwestern Medical Center in Dallas. The study was supported by funding from the National Institute of Neurological Disorders and Stroke, the National Institute on Aging, and the TAO Consortium by the Rainwater Charitable Foundation.