A psychedelic picture of the Parkinson's shuffle

Physicists use fractal analysis to help monitor the
disease's progression. Gaiting it right

By Robin Searle

Efforts to chart the course of Parkinson's disease by analyzing walking patterns have been depressingly unsuccessful up till now. At last the picture is starting to brighten. The welcome change comes from the design of an easily worn device called a triaxial accelerometer that measures gait. The data is then studied using fractal analysis.

The novel approach to gait analysis could eventually enable clinicians to easily and accurately follow the disease's progression and come up with more relevant treatments. A sensor is placed on the patient's lumbar region, which measures movements in three directions (vertically and in side-to-side and front-to-back horizontal planes). The data is collected by the triaxial accelerometer, along with information collected at the same time on walking speed, and stored in a database.

That's where the fractal analysis comes in. A fractal pattern maintains its same geometry when viewed close-up and at a distance. Examples include clouds, tree branches, a mountain range, a rugged maritime coastline, and a heartbeat. The researchers used the maximum likelihood estimator (MLE) based fractal analysis method to break the irregular walking patterns of Parkinson's patients into simpler, analyzable chunks of data. Analysis of these data components produced a "fractal measure" score of 1 to 2. The higher the score, the more complex was the body motion.

For Parkinson's disease patients, the fractal average of 1.48 was higher than that of healthy elderly patients (about 1.3). Their complex gait was accompanied by a significantly slower walking speed — means of 0.73 for Parkinson's disease and 1.25 for healthy subjects.

"The correlation between the complexity of body motion and walking speed for Parkinson's patients shows a drastic difference between them and healthy elderly subjects," conclude the authors. The more complex Parkinson's motion did not result just from an altered walking speed. Rather, say the authors, "healthy elderly subjects can walk freely at their own speed since they preserve sufficient ability to control smooth movement."

The system was developed by the research team from two centres in Japan and colleagues in New Hampshire. The research is published in the inaugural issue of the new Institute of Science publication, Journal of Neural Engineering, in February.

The authors have plans "to use the MLE-based fractal analysis method to assess the effectiveness of prescribed treatments and therapies."