Aug 14 2014

An App To Monitor Parkinson’s Disease

In 2000 Michael J. Fox began a non-profit organization to support research into Parkinson’s Disease (PD). This was shortly after he was diagnosed with the disease. Since then Fox has been the model celebrity spokesperson and advocate. (He doesn’t kibitz, he just raises awareness and supports the science.)

Now his foundation, together with Intel, have developed a wearable device and accompanying app that can monitor the symptoms of PD in real time 24 hours a day. This is an interesting application of technology, and something that we are beginning to see more, and will likely increase in future.

PD is a neurodegenerative disease affecting a part of the brain called the substantia nigra. Neurons in that structure produce and release dopamine. These neurons are part of a circuit (the extrapyramidal system) that essentially monitors and adjusts the sensitivity or gain of the motor system. It’s a sensitive feedback loop that keeps our movement smooth. If the gain is turned up too high then we would constantly be moving and writhing. If it’s turned down too low, then we start to freeze. People with PD have the gain turned down too low.

The symptoms of PD are stiffness, decreased movement, tremors, and difficulty walking and maintaining posture. The new device is worn like a watch, and it can monitor movements with 300 data points per second. This generates about a gigabyte of data per day, which can be uploaded to a database using a smartphone app.

This massive amount of information can potentially be very useful to researchers. They will be able to map fluctuations in the control of PD symptoms throughout the day. This could be used in drug trials to record the effect of PD medication. It can be used to mine for environmental factors that affect symptoms, such as sleep, eating, and time of day.

While we have some clues, we essentially don’t know what causes PD and we don’t know how to cure it. Treatment is therefore mainly symptomatic. It’s a delicate system, and it becomes increasingly difficult to obtain good control of symptoms as the disease progresses.

Therefore, in advanced PD, making small tweaks or improvements to the treatment regimen can have a significant impact on the quality of life of PD patients.

I also love the idea of continuous real-time ambulatory monitoring of medical parameters. One of the frustrations of practicing medicine is that we often have incomplete and subjective information about what is happening with patients. We spot check things like blood pressure, laboratory values, neurological function, whatever, and otherwise rely on patient report to document how they have been doing.

There are specific tests that involve monitoring a parameter for 24 hours or so, like a Holter monitor which is essentially a 24 hour EKG.

Portable electronics and smartphones, however, open up the possibility of gathering massive amounts of objective data over long periods of time, data that can then be analyzed by computer to look for trends and anomalies.

If used correctly, this type of data can provide very useful feedback with which to adjust treatments. It can also be used diagnostically.

There is a potential downside, however. It’s easy to be overly impressed with data, and to read significance into noise. I predict that as such systems come into legitimate use, they will also be abused by marginal or fringe practitioners. Small and insignificant fluctuations will be used to give bogus diagnoses, or to demonstrate response to bogus treatments. Large sets of data like this are ripe for mining data, confirmation bias, and creating the illusion of legitimate science.

Conclusion

This new device developed by Michael J. Foxes charity and Intel looks very promising and I would predict that it will find its uses in the management of PD patients.

The technology also is part of a trend toward real time monitoring of medical parameters generating massive amounts of data. This is an exciting development, as there are many opportunities that clever researchers can exploit.

Big data comes with a caution, however, as it is also ripe for exploitation by the pseudoscientific or unscrupulous.

Share

7 responses so far

7 Responses to “An App To Monitor Parkinson’s Disease”

  1. SteveAon 14 Aug 2014 at 9:44 am

    I’m guessing this is the type of continuous medical monitoring that the iWatch and its digital brethren are rumoured to be capable of (now, or in the future).

    A colleague of mine recently suffered a serious heart attack. I wonder if this kind of 24/7 surveillance would have given him some warning?

  2. Bruceon 14 Aug 2014 at 9:59 am

    As wonderful as somethign like this is:

    “It’s easy to be overly impressed with data, and to read significance into noise.”

    This was my initial thought and following on from that and what SteveA said; I would be very wary of a device that was hooked up to a population 24/7 that would potentially result in false alarm emergency service call outs unless you had very strict parameters that could not easily be mimicked. The cost of the almost inevitable false alarms would need to be weighed very carefully with the benefits (And I am not just talking financial costs, but emergency services being called out to false alarms reduces their ability to respond to real emergencies).

    Of course, it would also depend on how early the warning was, if you can stop a potential emergency service callout with a simple visit to a medical practitioner or by taking pre-emptive medication, you might be on to something.

  3. Bill Openthalton 14 Aug 2014 at 12:10 pm

    I agree there is a potential for abuse by the unscrupulous, but one of the problems we have investigating many chronic diseases is lack of data, be it for the efficacy of blood pressure medication, blood sugar level in diabetics, or the onset of convulsions in young children. The ability to monitor 300 data points per second, and a framework to collect the data for analysis should be of tremendous benefit for many.

    One more reason to like Michael J Fox.

  4. Ori Vandewalleon 14 Aug 2014 at 1:19 pm

    The efficacy of a heart attack early warning system would vary based on the type of warning it gave. For example, if it warned the patient with bright blinking red lights and alarm bells, it would probably never be wrong.

  5. Sylakon 14 Aug 2014 at 11:07 pm

    This sound really good, could suspect that this kind of Tech could also be adapted and used with other neurological diseases ? Like ALS Or MS ?

    Maybe even for others types diseases. A lot of medical conditions requires you to tell your doctor how you feel, how is you pain today etc, and sometimes it is hard, sometimes the patients himself does not really know. Symptoms can be hard to describe ( I know about that, I remember quite a few times having trouble describing to my doc or even my dentist my symptoms). But with that gadget, Doctors will have objective data.

    The large quality of data must be check carefully. It remind me when you talked about Overpowered studies on SGU. Like you points out, with so much data, some will find statistically significance for anything they wants. ( despite those “finding” not being clinically significant ) But every new tech and science can be used by cranks, Doctor and scientist working with it must be careful.

    Kudos to Monsieur Fox, I always like him. He also seen to view his disease with some humour. He was on a episode of “curb your enthusiasm”, there were jokes about PD, In the episode Larry david is stupid, as always, and Fox is super mad toward him, it is really funny.

    With all the the celebrities promoting non-sense and conspiracy theories ( Woody Harrelson? Why man? such a good actor!) it is good to see some promoting science and engineering. And Intel are also cool, I think they also made the tech for Hawking talking machine ( he select word of letter with his eyes i think something like that)

  6. SteveAon 15 Aug 2014 at 6:22 am

    Ori Vandewalle: “The efficacy of a heart attack early warning system would vary based on the type of warning it gave. For example, if it warned the patient with bright blinking red lights and alarm bells, it would probably never be wrong.”

    How about: “You are going to die in ten seconds…nine…eight…”?

  7. Pete Aon 16 Aug 2014 at 8:37 am

    SteveA, don’t forget the two-button dialog box:
    Accept
    Remind me nearer the time.

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.