Dr. Chih Lai Partners to Fight Parkinson's Disease
The patient reaches for his glass of water on the bedside table, his hand trembling and his head shaking involuntarily. With a twitch of his hand, the glass tumbles and the water spills to his pillow. The nurse rushes to help. The patient keeps apologizing to the nurse and tries to move his rigid body away from the wet pillow. The patient has just been diagnosed with Parkinson’s disease, along with another 10 million worldwide.
Involuntary and rigid body movements are typical symptoms of Parkinson’s disease. Various treatments will be prescribed to Parkinson’s patients based on an evaluation of the severity of the movement disorder. During this test, patients are asked to perform a routine of passive capabilities. The test is very subjective, based upon rater experience and training. It provides a single snapshot of a patient’s condition, is only available when patients are in the clinic, and cannot be performed continuously or repetitively because of patient and rater fatigue. As a result, the medicine that is prescribed is based on a subjective snapshot evaluation and it cannot be dynamically adjusted based on a patient’s conditions.
With its great potential to help Parkinson’s patients, a video motion analysis project was initiated by Dr. Dwight Nelson, a former St. Thomas biology professor who now works as a chief scientist in the Neuromodulation division at Medtronic. Requiring an interdisciplinary collaboration, it needed neuro-scientists to clearly define research goals and interpret the findings. It also needed to examine enough patients with various movement disorders over a long period of time to statistically verify the validity of final findings. Dr. Aviva Abosch, from the Department of Neurosurgery at the University of Minnesota, provided videos for this study. Dr. Chih Lai, St. Thomas computer scientist was invited by Nelson in April 2012 to participate in this project since he had many years of video motion analysis and data mining experience.
A challenge of this project is the ability to quantify the movement disorder so the numeric results can be used by healthcare providers to diagnose, treat or further movement analysis. There was a need to develop a method that could quantitatively describe irregular head or body swing (known as dyskinesia). Dr. Lai created an algorithm that has the capability of measuring tortuosity motion similar to measuring the curvature of a river.
The methods developed by Lai have been applied to analyze six patients with different movement disorders, each with over 38 hours of continuous analysis. The accuracy was evaluated at 71 percent to 97 percent with over 500 GB videos. As a result of this project, two patents were filed in summer 2013. The research team also hope that one day this technology can be integrated with video-conferencing so physicians can provide telemedicine and diagnosis to the Parkinson patients living in rural areas.