Home > Neurology > Computational model of Alzheimer’s links symptoms, brain anatomy, mental processing

Computational model of Alzheimer’s links symptoms, brain anatomy, mental processing

Journal
Nature Communications
Reuters Health - 05/04/2022 - A novel conceptual model of Alzheimer's disease, developed using functional imaging, connects clinical symptoms to brain anatomy and function, researchers say. 

The model compresses complex brain anatomy relevant to dementia symptoms into a conceptual, color-coded framework that shows areas of the brain associated with neurodegenerative disorders and mental functions. Imaging patterns shown in the model relate to the symptoms experienced by patients (e.g., memory, social cognition, visuospatial processing). 

"Our study showed that there is a simple relationship between mental functions and brain anatomy that can predict many aspects of Alzheimer's physiology, including the severity of tau protein pathology found at autopsy," Dr. David Jones of Mayo Clinic in Rochester, Minnesota told Reuters Health by email. 

"It is also able to relate this simple brain-behavior mapping to large-scale brain networks, mental task activation patterns, and a diverse array of clinical syndromes related to impaired mental functioning," he said. 

The model also suggests, he added, that "the relationship between mental functions and brain anatomy relevant for the medical care of Alzheimer's disease and related conditions... only requires understanding a handful of patterns of organization across the brain and the general functions associated with them," rather than hundreds of brain regions or large-scale networks. 

"This approach is analogous to breaking down the problem of flight to the simple concepts of lift, thrust, and drag to guide engineering decisions about the more complex details of plane design," he said. 

As reported in Nature Communications, Dr. Jones and colleagues developed the model using F18-fluorodeoxyglucose PET (FDG-PET) performed on 423 cognitively impaired participants (median age, 77.4; 58%, men), plus various computational techniques. 

The predictive ability of the model for changes associated with Alzheimer's physiology was validated in 410 individuals (median age, 74.5; 54%, men) from the Alzheimer's Disease Neuroimaging Initiative. 

Additional validation was obtained by projecting a large amount of data from normal aging (1,131 participants) and seven age-associated dementia syndromes: typical Alzheimer's disease (137), dementia with Lewy bodies (72), behavioral variant of frontotemporal dementia (33), semantic dementia (11), posterior cortical atrophy (15), and logopenic variant of primary progressive aphasia (8). 

Fifty-one percent of the variance in glucose use patterns across the brains of all participants with dementia could be explained by only 10 patterns. Each patient had a unique combination of these 10 brain glucose patterns that related to the type of symptoms they experienced. 

Dr. Jones said, "This model can unite all dementia syndromes into a single framework at a scale of observation that is meaningful for clinical brain-behavior mapping. This should inform new taxonomies for degenerative dementia syndromes and the associated functional symptoms that define them." 

Dr. Jesus Gomar, Assistant Professor in the Institute of Molecular Medicine at the Feinstein Institutes for Medical Research in Manhasset, New York, commented on the study in an email to Reuters Health. "The conceptual framework for this model is promising since it combines information from clinical symptoms, cognitive functions and brain functional networks that can be applied to other neurodegenerative disorders," he said. 

However, he noted, "The determination of cognitive impairment is based on a global measure. It would interesting to understand how objective cognitive performance in the core domains impaired in Alzheimer's disease relates to the FDG-PET-derived functional networks of brain activity." 

Further, he said, "The accumulation of tau protein is closely related to cognitive impairment; therefore, future models will potentially incorporate patterns of tau brain distribution to infer brain networks of tau pathology and brain function, and their relationship to specific cognitive impairments. In addition, the incorporation of longitudinal data may improve the predictive power of these computational models." 

SOURCE: https://go.nature.com/3KfCHgG Nature Communications, online March 28, 2022. 

By Marilynn Larkin  



Posted on