Researchers use computer-imaging technique to study Alzheimer’s

A new computer-assisted analysis technique may better identify cellular damage associated with the onset of Alzheimer’s disease.

Researchers led by Min-Ying Su of the University of California, Irvine, used a computer mapping technique to take readings from gray and white matter in lobes of the brain that had not been reported before.

To test the technique, MR images taken from 13 elderly people with mild cognitive impairment (and a mean age of 74) were contrasted with 13 healthy control subjects of the same approximate age.

Shown are original anatomic MR images (top row) and segmented gray matter maps (bottom row) from a coronal section of a control subject (left), from a patient without severe atrophy (center), and from another with severe atrophy (right). Reprinted with permission of Radiology.

Cellular damage causes an increase in the apparent diffusion coefficient — a measurement used to study the mobility of water in the brain. A segmentation program was applied to obtain maps of the gray and white matter, and the images were co-registered to a brain template; the defined brain lobes can be mapped to each person’s images and further separated into gray- and white-matter regions using segmentation maps. Finally, these regions were mapped to apparent diffusion coefficient maps to obtain values from the gray and white matter in each brain region. Computer analyses were applied for segmentation and imageco-registration.

The research showed that patients with impairment had increased mobility of water in the brain, regions of atrophy and higher levels of water content. Elevated values also were found in the brain’s hippocampus, corpus callosum and temporal lobe gray matter. Increased values in the hippocampus heavily correlated with inferior scores on memory tests.

Early detection of brain degeneration by computer mapping could allow earlier therapeutic intervention and possibly slow the progress of dementia, according to researchers. The results appear in the October issue of the journal Radiology.