Combined with other clinical and neurospychological evaluations, automated segmentation of the hippocampus on MR images can contribute to a more accurate diagnosis of Alzheimer's disease, according to a new study published in the July issue of Radiology.
In Alzheimer's disease, nerve cell death and tissue loss cause all areas of the brain, especially the hippocampus region, to shrink. MRI with high spatial resolution allows radiologists to visualize subtle anatomic changes in the brain that signal atrophy, however, segmentation is a complicated, lengthy process, according to researchers.
"Visually evaluating the atrophy of the hippocampus is not only difficult and prone to subjectivity, it is time-consuming," explained the study's lead author, Olivier Colliot, PhD, from the Cognitive Neuroscience and Brain Imaging Laboratory in Paris. "As a result, it hasn't become part of clinical routine."
Using an automated segmentation process with computer software developed in their laboratory by Marie Chupin, PhD, researchers measured the volume of the hippocampus in 25 patients with Alzheimer's disease, 24 patients with mild cognitive impairment and 25 healthy older adults. The MRI volume measurements were then compared with those reported in studies of similar patient groups using the visual, or manual, segmentation method, according to the study.
The results showed a significant reduction in hippocampal volume in both the Alzheimer's and cognitively impaired patients when compared to the healthy adults. Alzheimer's patients and those with mild cognitive impairment had an average volume loss in the hippocampus of 32 percent and 19 percent, respectively. Studies using manual segmentation methods have reported similar results, the researchers noted.
"The performance of automated segmentation is not only similar to that of the manual method, it is much faster," Colliot concluded. "It can be performed within a few minutes versus an hour.”