Volumetric MRI produces earlywarning on Alzheimer's disease

MRI measures of regional brain atrophy, taken at the University of California, San Diego, have established the basis for a new way to diagnose Alzheimer's disease.

Radiology professors Andres Dale, Ph.D., and Dr. Carl Hoh, neurologist Dr. James Brewer, and radiology resident Dr. David Karow saw that volumetric MRI potentially fulfilled the need for a less expensive and less invasive test for early diagnosis of the disease and for drug research and development.

Their suspicions were confirmed in a study involving MR and FDG-PET scans performed on 80 normal subjects, 69 subjects with single-domain mild cognitive impairment limited to memory loss, 156 subjects with MCI, and 68 mild Alzheimer's disease patients.

Subjects underwent whole-brain FDG-PET and a noncontrast T1- weighted MRI exam to optimize graywhite brain matter contrast. Data from both exams were processed through FreeSurfer, 3D reconstruction and segmentation software that assesses average differences in volume, thickness, and metabolic activity for 45 brain regions. The results were compared with brain segmentations from agematched normal subjects.

The study found that regional brain atrophy grew more severe as subjects progressed through single-domain MCI to MCI and early Alzheimer's disease.

Compared with normal older adults, single-domain MCI subjects had an average 9.5% reduction in hippocampal size, 6.2% less entorhinal cortex thickness, 5.5% less amygdala volume, and 4.1% less parahippocampal cortex. The morphometric losses were larger for subjects with MCI and larger still in the brains of mild Alzheimer's disease patients.

Atrophic changes observed with MRI were more conspicuous than metabolic differences measured with FDG-PET, Karow said.

“This suggests to us that if you are going to pick a region to diagnose single-domain MCI, MCI, or early Alzheimer's disease, measuring hippocampal volume is the way to go,” he said.

The study also confirmed the underlying premise for a reimbursable, MRI-based MCI/Alzheimer's diagnostic aid developed and applied clinically at UCSD.

Using NeuroQuant, a commercial version of FreeSurfer software, they are able to compare the volume of the patient's hippocampus and temporal horn of the lateral ventricle with the volumes of those regions in normal subjects and generate an age-related atrophy report. Alzheimer's patients typically have a hippocampal volume that is at least two statistical standard deviations below normal and a temporal horn of the lateral ventricle volume that is at least two standard deviations above normal, Karow said.