- Brian Cohn, Lead Scientist
- Ali Marjaninejad, Data Scientist
- Dilan Shah, Programmer/Project Manager
- Martin Shapiro, Health Strategist
- Serhan Ulkumen, 3D Designer/Programmer
Parkinson’s disease is a slowly progressing neurodegenerative disorder with symptoms including tremor, limb rigidity, slowness of movements, and problems with balance. Advancement of the disease can severely impact quality of life from physical disability to depression. We developed Insight, a VR patient-centric platform for Parkinson’s disease patients and their family. Insight empowers Parkinson’s patients to live a life of their choosing rather than that of their disease.
Insight augments Parkinson’s patient’s quality of life by integrating into their existing clinical care our VR assessment, management, and health education. Patients typically see a physician to monitor disease progression and adjust medication and rehabilitation therapy at set interval clinic visits based on symptoms. Insight stays with the patient throughout their life, continuously assessing the user between visits and aiding their providers at the clinic.
Our platform draws upon current care via third party health information such as medical records and movement data collected in VR to create an assessment of the patient’s health status, provide personalized rehabilitation exercises, and guide the physician team in data-driven decision making. Before the patient begins rehabilitation exercises, he/she touches the wind chime in the house which then transfers symptoms over to the virtual world. For the remainder of the experience, the wind chime movement and sound will signify the symptoms of the patient while the patient’s movements will now be tremor free. The patient will then be guided through evidence-based personalized rehabilitation exercises that while improving physical function will also collect data for disease progression assessment. At the end of the assessment, the patient receives an overview of their current health status including medications, Insight’s health score derived from symptom measurements, and an option to contact a physician through telemedicine. This physician will have a report generated by Insight that includes the collected symptom information.
Insight provides a platform for voluntary data collection and we have integrated neural network integration to allow for artificial intelligence deep learning of prognostic and treatment recommendations going forward.
The Insight Patient Data platform was built using a combination of Unity game engine and data analysis tools Matlab and Python.
The most crucial part of this project involved a transformation of the way someone with tremors actually moves versus how it appears they are moving when viewing their own hand through a VR device. Built by using a low pass filter or moving average, the SmoothedHandScript attached to the user's hand model captures transform position and rotation data from the VR tracked input object as input and outputs smoothed data for the transform of the hand model.
The inspiration for the environment was heavily influenced by one mentor, Hannah Luxenborg, who explained that rather than an art direction similar to a clinic, aim for a soothing ambiance.
A majority of the models were modeled and created in Maya by Serhan Ulkumen.
Using Unity's terrain system terrain was generated using a heightmap and trees were placed.
Essentially we collect X,Y,Z tremor values from the Transform of the VR controller and our data analysis provides details pertinent to patient and care-giver.
- HTC Vive
- For analysis we used: NumPy and Pandas
- For visualization: MatPlotLib
- Generated by our team
Thank you for visiting the repo. Please reach out to us if you are interested in learning more!