Home > Cool Tech > Microsoft Kinect camera helps assess symptoms in…

Microsoft Kinect camera helps assess symptoms in multiple sclerosis patients

Microsoft’s motion and depth-sensing Kinect camera may be most commonly used for gaming, but researchers at McGill University have come up with another use entirely for it: as a diagnostic tool to help evaluate the walking difficulties of patients with multiple sclerosis (MS).

The Kinect camera appealed to us because it was inexpensive, portable and easy-to-use,” McGill University postdoctoral fellow Farnood Gholami told Digital Trends. “We developed a framework which means that when a patient walks in front of the camera, our algorithm can identify whether the subject has gait abnormality or not — and also quantify that level of gait abnormality in terms of how serious it is.”

Related: Machine learning may be a key in helping diagnose Alzheimer’s earlier

At present, assessing these kind of walking difficulties — one of the symptoms of MS, caused by nerve damage as a result of the disease — is usually carried out by a physician. “The challenge with this kind of diagnosis is that it can be very subjective,” Gholami said. “Clinicians observe the walking of a patient and, based on their expertise, assign a clinical score. The problem is that different clinicians can assign different scores to the same subject. I had the idea of coming up with a more systematic and accurate way of assessing people with this kind of gait abnormality.”

To create their algorithm, the McGill University researchers first analyzed the movement of ten MS patients, along with ten members of a control group. By pinpointing certain gait characteristics, they were then able to come up with a tool able to distinguish between the walking patterns of people suffering from MS and those who were not.

“At this point, we’ve done the study to suggest this kind of framework can be useful,” Gholami concluded — noting that the work could also make it possible to carry out similar measurements for other diseases which cause gait abnormalities, such as Parkinson’s. “The next step is to come up with a product which can be used in clinics or even for remote monitoring on the part of the patient.

“Another possible application,” he continued, “would be using it in pharmaceutical companies, so that it can be used to analyze the effectiveness of certain medications in terms of their improvement on people’s walking. There’s more studying and improvement needed for the algorithm, but we’re certainly working towards creating a final product for use in the real world.”