Doctor's have used a subjective scale to track the progression of Parkinson's. People with Parkinson's typically visit their neurologist two to four times a year. This app makes communication between patients and their doctors easier by allowing self-tracking of a patient's disease.
What it does
Tracks the tremors and their volatility over a period of time. Graphs the results and allows you to submit the data to your health care provider after filling out an online form.
How I built it
Used a Leap Motion sensor to measure shaking of individual fingers in addition to the overall hand. The Processing language and various graphing API's were used to obtain and visualize the data in a user-friendly way.
Challenges I ran into
The Leap-Motion has lots of noise that could be mistaken for tremors. We needed to devise an algorithm to mitigate the noise levels and track only the important movements.
Accomplishments that I'm proud of
The algorithm to extract the volatility over time borrowed techniques from analyzing stocks. The math turned out to be simple but not obvious in its implementation.
The user interface is attractive.
What I learned
Gathering precise data (within millimeters) is difficult on a sensor that markets for less than $100 USD. But we did it!
Parkinson's is a disease that is tracked over very long periods of time. There are not many consistent ways of tracking a patient's progression on a daily basis. It is a medical field that can really benefit from innovative software development.
What's next for Tremor Tracker
Continue the development of the web front end to allow the tracking of different kinds of symptoms of Parkinson's. Specifically, posture and facial tremors could be detected using a sophisticated camera such as a Kinect. The Leap Motion could further be used to perform grip tests.