Imagine if Angry Birds ran the accelerometer in the background to measure whether players (who opt in anonymously) displayed Parkinson's tremors. They could collect millions of data points a day.
Parkinson's Progression Market Initiative(PPMI) is a really big deal. However, even with great tools like Fox Trial Finder, it's very difficult and expensive to get people to participate in clinical trials like PPMI. An estimated 3 million people in the world have Parkinson's; but nearly 30 percent of all clinical trials fail to recruit a single subject, and 85 percent of clinical trials finish late due to recruitment troubles. Which is why Mr. Little's work has me so fired up. It's not just because he has helped to develop a voice biomarker for Parkinson's. Even better, he has figured out how to use a phone - something that nearly everyone carries with them at all times - to create a cheap and easy way for anyone, anywhere to actively participate in the quest for a cure. As a mobile app developer myself, this has got me thinking of other ways we could use the other features of modern smart phones to efficiently crowd-source Parkinson's research. Here are some ideas - some of which are already being explored.
1. Using the phone's accelerometer to detect tremors. In 2011, researchers at UCLA published a study in which they used an iPhone 3G's accelerometer to detect Parkinson's tremors. The study was noted as a worthy proof of concept. And today's iPhones have way more sensitive gyroscopes, which could potentially read tremors even more accurately. They could collect millions of data points a day - from both healthy people, and people that self-identify as having Parkinson's - to perhaps develop a tremor biomarker for the disease.
2. Using the phone's camera to detect hypokinesia. Today's smart phone cameras are so sensitive that they can be used to measure a person's heartbeat. And modern facial recognition software is so sophisticated that it can detect a fake smile better than a human can. So, imagine if millions of people opted in to have Instagram anonymously submit their photos to a server running this kind of facial recognition software. If people self-identify as to whether or not they have Parkinson's, then the program could perhaps develop an early detection biomarker for hypokinesia-- the frozen or stony face that many people with Parkinson's display.
3. Using the phone's search engine, GPS and mobile payment systems to develop environmental/lifestyle markers. Google has a tool, called Google Flu Trends, that can detect a flu outbreak based on the types of search queries it receives. For instance, if a lot of people in a particular area start searching for "headache" or "cough syrup", then Google knows there may be a flu outbreak there. Clearly Parkinson's doesn't "outbreak" in a way that would cause a temporal spike in search terms. And clearly, search browsers are not limited to mobile phones. But, mobile phone searches might offer other interesting data because people can do them on the fly.
If people self-identified as to whether or not they have Parkinson's, it's possible that we could develop a "search biomarker" for the disease. The same could also be said for using a phone's GPS to identify environmental markers for Parkinson's, or the phone's mobile payment features (i.e. mobile wallets) to detect things like diet or lifestyle markers. The key to all of this is in having lots of people participate. But, as any app developer knows, this is surely possible if it's free and easy. Thanks, Mr. Little for helping to lead the way.