Inspiration

Tens of millions of individuals suffer from involuntary hand movement around the world. Even with such a significant impact on the general population, the effective diagnosis and differentiation of varying types of hand movement remain a challenge. There is no official test to differentiate between these types of movements. Currently, doctors (neurologists) use a combination of patient signs and symptoms, patient medical history, and a physical examination to come up with a diagnosis. In an effort to better diagnose and assess involuntary movement, we decided to automate the process with hopes of assisting doctors in the process of diagnosis and treatment assignment.

What it does

Our application uses Leap Motion technology to track the hand movement of users over a 10-second time span. Prospective patients are instructed to trace a straight line shown on the monitor. The program then returns summary statistics for the path the user actually traced, such as average amplitude, the standard deviation of the amplitude, and the number of "spikes" over the 10-second time span. Using carefully gathered medical data, we then determine whether the user displays signs of one of the four following hand movements:
Steady hand movement Myoclonus Tremors Athetosis

How we built it

We used Leap Motion to detect the tip of a pointer finger moved as steadily as he/she could. The Leap Motion allowed to gather data and we used statistics and other data analytic techniques to predict possible forms of involuntary movement that a person was experiencing, using a carefully calculated heuristic. We then describe the various diseases that associate with each form of hand movement and possible ways to prevent or decrease the effects of the diseases.

Challenges we ran into

Initially, our team was having trouble gathering data from the Leap Motion technology. Coming in, we were all inexperienced with the technology and there was a subsequent steep learning curve. After that, things became a lot easier as we had the data from the Leap Motion to work with.

Accomplishments that we're proud of

The mathematics required to interpret the data were very sensitive and intense and they required us to take many factors into consideration. We think that with more research and time we could advance our formulae to improve the recognition of the pattern of the hand movements.

What we learned

We learned that the Leap Motion technology has many practical and advanced applications. The experiments we performed here with the Leap Motion are only the beginning.

What's next for Biopath

Now that we have a working model for diagnosis, there is no limit to where we can take this project. Biopath should be applied in some hospital and clinical settings to identify flaws in the model. This will allow us, the developers, to improve our methods of classification. Additionally, we could improve our search methods to provide patients with more accurate and comprehensive therapy plans.

Share this project:
×

Updates