About this project
The goal of OneRing is two-fold:
- Bring our wearable monitoring ring into production for usage at local Parkinson’s clinics where patients and doctors will be able to benefit from the technology.
- Increase awareness about the need for technology to assist Parkinson’s patients with their daily lives.
Your funding will allow us to test and use the device with more patients, as well as build a campaign around the need for technology to help those with Parkinson’s. The rewards are structured around both of these goals.
Here's how we made the video at home: https://medium.com/@utkarshtandon/onering-how-i-bootstrapped-my-kickstarter-video-on-a-high-schooler-s-budget-3077302ceb76
What does OneRing do?
Parkinson's is commonly known as a movement disorder caused by a shortage of dopamine. This deficiency results in abnormal nerve functioning, causing a loss in the ability to control body movements. These motor distortions cause significant problems for Parkinson's patients on a daily basis.
OneRing is a novel tool for monitoring Parkinson's disease. Its intelligent machine learning technology has been trained to model various Parkinson's movement patterns including dyskinesia, bradykinesia, and tremor in order to generate smart patient reports on a daily basis. These reports provide physicians with time-stamped analytics about their patient’s movement severity at each hour of the day; allowing them to better prescribe medications. With smarter medications dosages/timings, Parkinson's patients would get quicker relief from their motor symptoms.
Design Specifications and Costs
Each component of the OneRing hardware can be seen below with its respective costs.
Accurately designed to perfectly hold electrical components and still be compact enough to be worn on one's finger.
Deployment of OneRing to a local institute will require the production of 10-15 rings in order to install rings of all sizes in each screening room. We will also provide one iPod Touch 6th Generation ($190) in each room (no more than 5 rooms will be supplied with rings/iPods) so that exams can be completed during regular appointments. Rings will also be sent home with patients who would like to use the device on a daily basis.
Prototypes and Previous Designs
OneRing started as a simple wrap around on someone's hand in order to validate algorithms. Over the past year several prototypes have been made to improve functionality.
We are developing a "one size fits all" ring using flexible plastic material, but due to the higher cost and slower production time we are sticking to our current design for the time being.
The iPhone App
In order to capture movement data from the OneRing hardware and provide a way for patients to begin and end their daily exams, we created the iOS app OneRing - Artificial Intelligence for Parkinson's Disease.
OneRing's hardware connects to the iOS app through low bluetooth energy technology. Collected data is analyzed by the backend machine learning algorithms and reports are returned back to the user in the app.
A Parkinson's patient can connect to their ring, start their day, end their day, view their ring's battery, and view their most recent report in one screen. By clicking view reports, the patient can see all their daily reports with datestamps, timestamps, and hourly movement severities. Each of these reports are sent to the patient's physician in order to accurately prescribe medications.
Machine Learning Algorithms
The core research behind OneRing lies in the developed machine learning algorithms. A statistical training model called Random Forest was used to automate decision making of movement severity classification. Features extrapolated from signal processing algorithms developed in Python are fed into the RF module to locate trends seen in training data. Each artifact in the below graphs represents a single feature vector produced by the various developed signal processing algorithms. These artifacts are used to train the final predication model which is an automated stage classification system similar to that of the Hoehn and Yahr scale used by physicians in current diagnostic measures.
The machine learning study could not have been completed without the publicly available dataset provided by the Michael J. Fox Foundation for Parkinson's Disease. The training data encompassed passively collected acceleration data for several Parkinson's patients of varying severities.
How It All Started and Looking Forward
I got the idea for OneRing when watching Muhammad Ali, who has Parkinson's disease, light the Olympic torch in 1996. I was not born yet to witness the actual event, but after watching the video on Youtube I wanted to learn more about the disease since I could see how strongly the movement disorder affected him in the video. I had additionally seen Back To The Future when I was young and my dad told me that the main actor, Michael J. Fox, has Parkinson's disease as well. This really inspired me to use my knowledge in programming to develop a device that could help Parkinson's patients deal with their motor symptoms.
Further Design Improvements
We look to make OneRing as simple and easy to use for a Parkinson's patient, and therefore ring size is a very important factor. Our current PCB is the only constraint towards making the ring smaller and therefore we plan to conduct miniaturization by cutting down on extraneous chips in order to support only Bluetooth connectivity and tri-axial acceleration. Your funding will help us achieve this peripheral goal, as we not only hope to bring our current design to Parkinson's patients, but also to create the best product possible.
Risks and challenges
Since the device is already built and has been undergoing testing for several weeks, we hope to encounter no problems and get rewards to backers by the estimated delivery. Nevertheless, I am a student so unforeseen delays could occur in the longer-term production, but the anticipated quantity for rewards is already met - so any delays for backers is an incredibly unlikely scenario.Learn about accountability on Kickstarter
Have a question? If the info above doesn't help, you can ask the project creator directly.
Support this project
- (36 days)