Software tool could produce early Alzheimer's diagnosis

After so far failing to concoct a viable treatment for Alzheimer's that can counter the disease in its late stages, researchers now think they can help if they catch it on the early side, which makes reliable diagnostics even more crucial to their work. A major research institute in Finland is touting a new software tool to accomplish this, something that could help at least some patients with memory problems gain a diagnosis at least a year before current technology allows. Systematic mathematical modeling, they believe, could also lead to far earlier and more successful treatment.

PredictAD is the software tool involved here. Designed by VTT Technical Research Centre of Finland (a nonprofit agency under state supervision), the tool was initially developed to help aid clinical decision-making. But in conjunction with clinicians at the University of Eastern Finland and Copenhagen University Hospital Rigshospitalet, PredictAD had some reliability in diagnosing Alzheimer's by processing patient measurements. Essentially, clinicians can wield this software tool to come up with a patient's diagnosis by comparing his or her memory measurements/imaging data with those from other patients in large databases. The end result: an index and graphical representation reflecting a patient's overall state.

For the study, they looked at records of 288 patients with memory problems, almost half of whom faced an Alzheimer's diagnosis 21 months on average after the initial measurements. Half of those patients could have gotten a definitive diagnosis at least a year sooner through use of VTT's software program/mathematical model.

How would this help? From a patient perspective, it is mixed. Early diagnosis doesn't mean a cure is possible yet, not by a long shot. But identifying Alzheimer's at an earlier stage gives drug developers a shot at testing new treatments for patients before their condition has become too advanced. VTT determined that its system is as accurate as a clinical diagnosis, and so its software tool could play an important role in enabling drug testing for patients before the disease has advanced too far for treatment to make a difference.

Other scientists are also focusing on identifying Alzheimer's at an early stage. Folks at the University of New South Wales in Australia, for example, believe that they have a blood test that will ultimately identify the disease before patients develop symptoms by linking early Alzheimer's onset to the number of proteins that transport cholesterol in the blood. VTT is also working with GE Healthcare to develop a serum biomarker that could predict Alzheimer's long before symptoms become apparent.

Plans call for testing the method further at a number of memory clinics in Europe. For more details on the scientists' initial work with their tool, check out the Journal of Alzheimer's Disease.

- read the release

Related Articles:
Alzheimer's blood test moves closer
15-minute test could tag Alzheimer's