Researchers using conventional, as opposed to high powered MRI equipment, and computer-
based methods to analyze the scans, are breaking new ground in the diagnosis of the telltale signs of
Alzheimer’s Disease, increasing the likelihood that there will be reliable ways of diagnosing the disease
non-invasively and before it is too late to do something about it.
Three studies, one on successful use of conventional MRI to image brain plaques in animals, and two
on using computers to analyze MRI images, are being presented at the Alzheimer’s Association’s
2008 International Conference on Alzheimer’s Disease (ICAD 2008), which is taking place in Chicago
from 26th to 31st July.
Autopsy exams that find amyloid plaques (abnormal protein deposits around brain cells) and other
characteristic lesions in the brain (such as neurofibrillary tangles, where unwanted strands of tau
protein collect inside brain cells) are currently the only definitive diagnosis of Alzheimer’s, so any
approaches that help to spot these signs non-invasively, while the patient is alive, and early in the
disease, is a huge step forward.
High powered MRI scanners have already been used to show images of amyloid plaques in animals,
as have PET scans combined with chemical markers, but until these studies, being able to image the
plaques using conventional, clinical strength MRI scanners, was not possible.
Conventional MRI scanners are more common and less expensive than other imaging machines, and
they don’t expose people to radiation compared to high energy technologies.
Vice president of Medical and Scientific Relations at the Alzheimer’s Association, Dr William Thies,
“As we get closer to the development of therapies that can slow or even stop the progression of
Alzheimer’s, earlier detection of the disease becomes crucial for early intervention.”
“Early evaluation and diagnosis is also important because some Alzheimer’s-like symptoms can be
reversed if they are caused by treatable conditions, such as depression, drug interaction, or thyroid
problems,” he added.
Early diagnosis also opens a host of other possibilities, such as access to the right medical skills,
medications, support programs, and so on. “Plus, there is the opportunity to participate in studies of
experimental drugs or other disease modifying treatments,” said Thies.
Thies said we may soon be able to use the approaches described in the studies to find out who is at
greater risk of Alzheimer’s.
Study Number 1: Amyloid Plaques in Rabbits
Researchers at Robarts Research Institute and University of Western Ontario, London, Ontario,
Canada fed laboratory rabbits a high cholesterol diet for more than 2 years so their brains formed
amyloid plaques. They then used clinical strength MRI scanners to take brain images from the
animals. These showed “signal voids”, or “black spots”, in several parts of the brain, including the
hippocampus, which is very important for memory. Autopsies showed that these areas had small
clusters of amyloid plaques together with high levels of iron, which the researchers suggest caused the
signal voids in the MRI images. Animals fed on a normal diet did not show these signal
One of the researchers, John Ronald, who is doing a PhD in Medical Biophysics at Robarts,
“Although some of the technology used to generate these images was designed specifically for rabbits,
this preliminary discovery hints at the promise of using clinical MRI scanners to visualize plaques in
people with Alzheimer’s.”
“Extension of these technologies to living animals is practical, and should allow us to study the course
of Alzheimer’s in animals over time,” he added, explaining that they customized the MRI equipment in
some very important ways, such as adding hardware that sees things smaller than 50 microns across
(0.05 mm), and is able to detect iron more sensitively.
Study Number 2: Developing an Alzheimer’s Disease Severity “Score” Based on Computer
Analysis of MRI Images of Tangles
As well as amyloid plaques, neurofibrillary tangles of tau protein that collect inside brain cells are
another telltale sign of Alzheimer’s, and there is a way of measuring these as an index of disease severity (post mortem)
using a “gold standard” called Braak staging.
Researchers at the Mayo Clinic, Rochester, Minnesota, developed a computer algorithm that takes
characteristic information from a person’s three-dimensional MRI scan and compares it to a standard
score called STAND (STructural Abnormality iNDex, a measure of tangle severity), which was
calibrated from comparing people’s MRI scans before death with their Braak stage after death.
The person is given a STAND score based on the degree of atrophy in their brain compared to atrophy
patterns from a library of MRI scans of 160 Alzheimer’s and 160 patients who did not have the disease
(cognitively normal). A positive STAND score means more Alzheimer’s-like, and a negative score
means more normal-like. STAND scores are adjusted for demographcs, and have a 90 per cent
accuracy, said the researchers, who validated the adjusted STAND scores against Braak staging
using 101 patients who had MRI scans within 4 years of their death and who underwent post mortem
One of the researchers, Dr Prashanthi Vemuri, said:
“This study shows that information extracted from MRI scans can accurately capture the severity of
Alzheimer’s tangle pathology.”
“While this work needs to be replicated and confirmed, because there is evidence indicating that
structural changes in the brain precede cognitive symptoms, STAND scores may also prove to be
useful for early identification of Alzheimer’s,” added Vemuri.
Study Number 3: Measuring Alzheimer’s-like Brain Abnormality in Normal Elderly
In a previously unreported study, researchers at the University of Pennsylvania, Philadelphia, and
National Institute on Aging, Bethesda, Maryland, analyzed computer-based images (using a new
approach called advanced high-dimensional pattern classification) of MRI scans of participants in the
Alzheimer’s Disease Neuroimaging Initiative (ADNI) and developed an index of Alzheimer-like brain
Then, in a new study presented at the conference, for 9 years they studied 109 elderly participants who were
cognitively normal and a small group who had mild cognitive impairment (MCI), from the
Baltimore Longitudinal Study of Aging (BLSA). They found that Alzheimer’s-like patterns of brain
atrophy were more common in participants who were over 80 years old. The rate of progression,
based on their new index, was also higher among the older participants.
They found that healthy elderly participants, and those with MCI, where both had the Alzheimer’s-like
brain atrophy patterns, performed worse in memory tests, compared to people who did not have the
abnormal brain patterns and had lower Alzheimer’s-like index values.
One of the researchers, Dr Christos Davatzikos, of the Department of Radiology at the University of
“Although the clinical significance of these Alzheimer’s-like patterns of brain atrophy must be further
evaluated, we are very hopeful that these pattern analysis tools will provide early indicators of brain
changes that resemble those seen in people with Alzheimer’s, years before memory problems are
“Direct visualization of [beta]-amyloid plaques in hypercholesterolemic rabbits using clinical
“Antemortem MRI based structural abnormality index (STAND)-scores correlate with
postmortem Alzheimer disease Braak stages.”
“Longitudinal progression of Alzheimer’ disease-like patterns of brain atrophy in normal elderly
Christos Davatzikos and Susan Resnick.
Click here for ICAD
Source: Alzheimer’s Association.
: , PhD