Sandy Pentland on Reality Mining: Phoning In the Data

Professor Alex (Sandy) Pentland is the co-director of the Digital Life Consortium at the Massachusetts Institute of Technology and was co-founder of the Center for Future Health at the University of Rochester, which we write about often here on the blog. Pentland has a grant from Pioneer to explore the potential role of reality mining technology - a concept that he helped develop - in medicine and in public health.  We asked him to tell us about this work, and he responded: 

We live our lives in digital networks. We wake up in the morning, check our e-mail, make a quick phone call, commute to work, buy lunch. Many of these transactions leave digital breadcrumbs – tiny records of our daily experiences. Reality mining, which pulls together these crumbs using statistical analysis and machine learning methods, offers an increasingly comprehensive picture of our lives, both individually and collectively, with the potential of transforming our understanding of ourselves, our organizations, and our society in a fashion that was barely conceivable just a few years ago. It is for this reason that reality mining was recently identified by Technology Review as one of “10 emerging technologies that could change the world.”

As pointed out in a recent Nature article, the single most important source of reality mining data is the ubiquitous mobile phone. Every time a person uses a mobile phone, a few bits of information can be collected. The phone pings the nearest mobile-phone towers, revealing its location. Accelerometers already in some phones can record patterns of physical activity, and the phone’s signal processing hardware can analyze the user’s speaking patterns. With the aid of data-mining algorithms, these data could shed light on the user’s health behaviors, creating new ways of improving their health.

 

To illustrate, consider two examples of how reality mining may benefit individual health care. By taking advantage of special sensors in mobile phones, such as the microphone or the accelerometers built into newer devices like Apple’s iPhone, important diagnostic data can be captured. Commercial trials by start-up Cogito Health are demonstrating that we can accurately screen for depression from the way a person talks -- depressed people tend to speak more slowly, a change that speech analysis software on a phone might recognize more readily than friends or family do. Similarly, experiments in my laboratory have shown that monitoring a phone’s motion sensors can also reveal small changes in gait, which could be an early indicator of ailments such as Parkinson’s disease.

 

Perhaps the greatest potential of reality mining of mobile phone data is to create a personalized health system (as opposed to a heathcare system): a set of information tools that helps people thrive, staying healthy and happy during their entire lives. Such a system would be owned by the individuals themselves, not by hospitals or clinics.

 

Best Buy, CVS, and Wal-Mart are already queuing up to sell and service the tools such as these that allow people to manage their health. The vision is that is emerging is of a health system built around mobile phones with special sensing capabilities to record your daily and weekly patterns, smart bathrooms that keep track of new types of vital signs, smart exercise equipment that knows your personal patterns, and more…all provided by consumer electronics and similar industries. By building a health system that supports lifelong health, we can make sure our healthcare system is used in the most efficient way. And, even more importantly, we can help citizens of the United States of America achieve far more healthy, happy, and even thriving lives.

 

photo credit: Julien Pacaud. 

Comments

[stephen texeira]

Professor Pentland has some interesting ideas, but they are far from the reality of most care givers in this country. The idea that providers will be accessing data about their patients gathered from their cell phones so that they can provide more proactive and preventive care is frankly absurd given the fact that most doctors in this country still view email as a trendy fad. There is also the fact that there are literally hundreds of forms of preventive medicine that are much cheaper, more immediately available and would have far greater positive outcomes. Why don't we spend a few more dollars on the low-tech solutions we have that are not being implemented before we start getting all fancy.

[Al Shar]

While there is no doubt that there are many inexpensive forms of preventive medicine (most of them ignored by those who have the most to gain from them), I think that Stephen Texeira's post misses the point at three levels. First, the tools are meant mainly for the individual using the cell phone, not the provider. This is an important change - the patient is at the center and in control, not the provider. Second, reality mining is a way of efficiently analyzing complex and different data that can be communicated to professionals in a more succinct way, saving time and providing more actionable information. Finally, the cell phone is getting to be close to universal, inexpensive and "smarter". While it's true that devices like the iPhone are expensive, there's lots that can be done on some basic devices. One arm of Dr. Katz's diabetes test uses a basic phone and an open source tool.

[Sandy Pentland]

One additional area where reality mining will have real impact is public health, especially on behavior change. The ability to communicate to particular social groups allows us to harness peer-to-peer social
effects for behavior change. I expect to see major research progress in this area during the next few years.

I am also seeing many large consumer companies making big plans for home products that harness reality mining techniques to provide a "health dashboard" to the consumer (and doctor). There is widespread agreement at the very top levels of these companies that there is a major commerical opportunity in creating a consumer health (not health care) system.

[Charles Beauchamp MD, PhD]

Personalized medicine using cell phone technology is certainly an elitist way to go. A more proletarian way to go is what the VA does with its PHR, MyHealthyVet. Here the patient can view data and informtion, request refills for medications on line and eventually enter one's own personal comments. Link this to smartly annotated Problem Lists as the VA already has and interdisciplinary and patient-generated interventions and you have a low cost and presumably (remains to be evaluated) cost-effective health care system focused more on personal outcomes than on medical processes with copious documentation for better reimbursement. Personal Medicine as described will fit very well into the current way of doing business. A cooperative and evolving EMR/EHR/PHR that the taxpayer has already paid for and has demonstrated its cost effectiveness in the
VA is definitely the way to go, in my opinion. The RWJF consistently overlooks this fact and consistently finances attempts to reinvent this wheel. The RWJF needs to disrupt its own thought processes an realize that there are organizations that have done a lot to cut through the Gordian Knot of health care reform and progress in the direction of pragmatic personal medicine.

[Neal Kaufman]

In my opinion for reality mining to have its transformative impact it must be linked into an healthcare delivery approach that provides patients with the day-to-day support they need to prevent disease and to self-manage their conditions if they are ill.

In the connected era that means just in time delivery of the personalized and up-to-date data and information a person needs to have the knowledge and skills to make wise choices.

It means supporting patients to easily and accurately keep track of their performance and use that knowledge to plan and implement new approaches to reaching their goals.

It means providing tailored messages and experience that speak to each person based on their unique characteristics, their performance on key behaviors and their needs at that moment in time.

It means helping patients link directly to family and friends for critical support, and link to their many providers to help integrate medical care with everyday life.

Just giving them the data and information they need isn't enough...it should be linked to the support they need and deserve.

[Sandy Pentland]

I belive that Dr. Beauchamp has the access issue exactly backwards: poor people have cell phones, not PCs, and increasingly this is their main portal to the web.

This, however, completely misses the main point: Health support requires data about the individual's day-to-day experience... things like, how much are they socializing? How strong is their support group? How much are they moving about? What is their weight variation? How do they sound when they talk (e.g., voice analytics), and all those other things that go into clinical global impressions and quality of life assessment.

The VA system doesn't have a way to continuously collect this sort of data, and so it is restricted to basing feedback on in-clinic notes and self-report data. Health support based on this very thin and distorted view of the persons life leaves much to be desired.

To me Dr. Kaufman's point is the correct one: we need to focus not only on collecting *continuous* data relevant to clinical impressions and quality of life, but most importantly on turning this raw data into meaningful feedback and an effective system of health support.

[moncler outlets]

Nice, and thanks for sharing this info with us.Good Luck!