Early ID for AD

New software for the automated analysis of MRI brain scans could help specialists identify cases of mild cognitive impairment years before full-blown Alzheimer's disease is apparent, according to US researchers.

The software developed by a team at the Martinos Center for Biomedical Imaging at Massachusetts General Hospital (MGH) is described in a forthcoming issue of the journal Brain. It works by accurately differentiating patients with mild cognitive impairment or Alzheimer's disease from normal elderly individuals based on anatomic differences in brain structures known to be affected by the disease.

"Traditionally Alzheimer's has been diagnosed based on a combination of factors - such as a neurologic exam, detailed medical history and written tests of cognitive functioning - with neuroimaging used primarily to rule out other diseases such as stroke or a brain tumour," explains Rahul Desikan of the Martinos Center and Boston University School of Medicine. "Our findings show the feasibility and importance of using automated, MRI-based neuroanatomic measures as a diagnostic marker for Alzheimer's disease."

According to the team, mild cognitive impairment is present in approximately one in five of elderly people, a figure that rises to forty percent in people over 85. Of these cases 80% will ultimately develop Alzheimer's within five or six years.

Drugs to slow the progression of Alzheimer's disease have been in development for many years, but to be optimally effective treatment in the earliest stages of the disease is necessary to delay progression to dementia. However, until now the necessary early diagnosis was difficult, if not impossible.

Desikan and colleagues Bruce Fischl, Nicholas Schmansky, Douglas Greve, and David Salat, working with Christopher Hess, William Dillon, Christine Glastonbury, and Michael Weiner, at the University of California at San Diego, Howard Cabral of the Boston University School of Public Health, and Randy Buckner, Harvard University and Howard Hughes Medical Institute, have looked to MRI to help them reveal early diagnostic markers. Now, using a computer program known as FreeSurfer. FreeSurfer is a set of automated tools for reconstructing the brain's cortical surface from structural MRI data. It can then overlay functional MRI data on to this reconstructed surface and allow medical researchers to examine a number of neuroanatomic regions across a range of patients.

In the first phase of the study, the investigators examined magnetic resonance images of almost 100 elderly individuals. Some of the people in the sample had been diagnosed as having mild cognitive impairment, the others were cognitively "normal". The team found that there were three brain regions that displayed structural differences between the normal controls and those individuals with mild cognitive impairment; accuracy was 91%.

The team explains that the same three regions of the brain - the hippocampus, entorhinal cortex and the supramarginal gyrus - have all been implicated in Alzheimer's disease in earlier pathological and imaging studies.

Having succeed in demonstrating 91% accuracy with this small test sample of participants, the team then turned to the Alzheimer's Disease Neuroimaging Database and analysed imaging data from 216 individuals. 94 of those patients were "normal", 58 had mild cognitive impairment at the time of imaging and went on to develop dementia, and 65 had probable Alzheimer's diseased based on their clinical symptoms. They also carried out a series of neuropsychological tests on the patients and obtained samples of cerebrospinal fluid when possible.

They had an even higher success rate with this sample than the initial tests. Automated MRI measures of the same three brain regions discriminated between individuals with mild cognitive impairment and normal elderly controls with 95%. Patients with Alzheimer's disease were identified from their brain scans using Freesurfer with 100% accuracy.

To validate their data, the team demonstrated a strong correlation between the clinical and cognitive tests for dementia, in particular decline in memory function, and with cellular biomarkers for the disease in the form of tau and amyloid proteins present in the cerebrospinal fluid samples.

"Our results indicate that these automated MRI measures are one effective way of identifying individuals in the earliest stages of Alzheimer's disease,2 says Desikan. However, he cautions that before the technology can be used in the clinic there are now required several follow-up studies.

"Those include determining whether these automated MRI measures can accurately predict which individuals with mild cognitive impairment will progress to Alzheimer's; seeing if they can differentiate Alzheimer's from other neurodegenerative diseases; assessing how these measures do at early diagnosis, compared to other measures such as cellular biomarkers; and then validating all of these findings against the gold standard for diagnosis, post mortem examination of brain tissue," he says.

Fundamentally, the results demonstrate that automated MRI measures can serve as a surrogate marker revealing underlying neuropathology and as a non-invasive method for identifying the early stages of Alzheimer's disease. 

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.