Skip to main content

American Diabetes Month 2015 - Day 7: Medical Technology Part 2

Day 7.  As I said before, I'll be continuing the theme of explaining the technology that I use with my devices, then go on to explain some of the holes.  WARNING: There is some fun mathematical discussion in this post, if it scares you, feel free to skip today's lesson.

Where we last left off (on Day 5) I had talked about meters and my insulin pump as far as technology goes.  Today, I would like to touch on glucose sensors.  I first tried out glucose sensors back in 2009 when I was in Iowa City with my Medtronic Paradigm 712 insulin pump.  At the time, the sensors were very large, in both size of needle being inserted and in the transmitter that attached to the inserted needle, but they also have a very short lifespan for being inserted, mostly due to their size.  I ended up stopping with the use of my continuous glucose monitoring for two major reasons.  One of those reasons is still an issue with CGM sensors now, but the issue with CGM sensors is that the first day of data is garbage.  The starting calibration of the sensors takes 2 glucose tests.  In the medtronic sensors (at the time) it took 2 finger sticks at a 2-hour interval.  With the Dexcom G4, it's 2 finger sticks in succession 2 hours after the initialization period.  Once the first step of calibration is completed, the only further requirement is an "update" calibration every 12 hours afterwards.  This is pretty easy to cover, but you can also do that calibration at 6 hours, or 4 hours, or really any time before the 12 hours.

The reason that the first day of data is garbage (at least the first 12 hours) is that you are taking 2 glucose measurements in a pretty short time period, and from what I can tell, the correlation between your blood glucose and interstitial glucose is correlated with a linear scale, i.e. y = mx+b is solved where you know y (your blood glucose) and x (your interstitial glucose) so you're solving for m (the slope) and b (y-intercept).  I'm sure with more glucose measurements, there is a certain amount of fit that is being looked for with the 2 most recent glucose measurements being taken as the main consideration for slope, but previous ones (based on a time going back) are slowly ignored.  This is good, because as sensors age under the skin, there is likely some degradation of the signal that is transmitted to the transmitter, thus causing measurements to go bad after some amount of time.  The first day is bad, because there's only 2 points, and likely a small amount of interstitial data to go off of, so the first slope calculation is likely more guessing with the slope and intercept than anything else.  If I know anything from the work that I've done at my job, it's likely a much more complex equation, but it has been simplified to work on low-processing-power devices (like insulin pumps or CGM receivers), though how hard can algebra really be?

This issue of being bad for the first 12-24 hours was a huge issue with the medtronic sensors as their rated time in the body was 3 days (same as the infusion sets).  which meant that during a week, there were at least 2 days that were not useful for getting data on the glucose.  With the Dexcom sensors, the devices that deal with the CGM sessions limit the sessions to 7 days, after which there is no longer any readings coming back from the transmitter to receiver.  This means that 6 out of 7 days a week are useful in determining the conformity to the better glucose values.  I think I'll have to continue this again.  Technology in Diabetes is definitely a topic that I'm very involved with, and I want others to know the technology that I use, and just how archaic it really is.

NOTE: I know that medtronic has changed its sensors since I used them back in 2009, but I have not been able to try them out.

Popular posts from this blog

Stories of Hacking: Part 1

So after my first post,Why Hack? I have decided to start a multi-part series on some of the things that really got me interested in programming.  Some of these things may kind of scare people who I gave assistance to in the past as I've done some things that may later in life make you a little bit upset.

Breaking Down the Cost of Diabetes #dblogweek

Today marks the second day of dblogweek for 2017 and today's topic is "The Cost of a Chronic Illness" which I feel like I've covered before, but it would appear it's only thinking of the cost that I've done, not writing about it.
Insulin and other diabetes medications and supplies can be costly.  Here in the US, insurance status and age (as in Medicare eligibility) can impact both the cost and coverage.  So today, let’s discuss how cost impacts our diabetes care.  Do you have advice to share?  For those outside the US, is cost a concern?  Are there other factors such as accessibility or education that cause barriers to your diabetes care?

Using Diabetes as a Guide #dblogweek

Today marks the final day of dblogweek for 2017 and today's topic is "More Than Diabetes" which is a topic focused on letting readers of dblog week get to know me better. Below is the prompt, makes pretty good sense.
Lets wrap up the week by sharing a little more about ourselves, beyond the chronic illness we or our loved ones live with.  Share an interest, hobby, passion, something that is YOU.  If you want to explore how it relates to or helps with diabetes you can.   Or let it be a part of you that is completely separate from diabetes, because there is more to life than just diabetes!  For people who have previously read my blog, you would likely know some of the story of me outside of diabetes, but after diagnosis (just before my senior year in High School) it really did shape my life quite a bit, it changed the course of my life forever. It started me on the course to go into a major in college where I could use my skills to make a difference in the lives of other p…