Marcus is cool.  I’m pretty certain about this because he has diabetes, thus making him part of a group of people I trust without justification outside of the whole “not making insulin, either” part.  (And he makes entertaining videos, but that’s a whole separate post.)  Over the next few weeks, I’ll be hosting some guest posts here and there, and today, Marcus is taking over SUM to talk about how his day job intersects with his all-the-time disease, and what kind of help he’s looking for from our diabetes community.

*  *   *

Like most of you, I have plenty of thoughts on what device manufacturers for diabetes should be doing differently. Through my thirty-two years of being T1, I’ve used countless pumps, CGMs, and meters. Opinions? I’ve got plenty of them.

I suppose that’s why I jumped at the chance to get an insider’s view as Chief Marketing Officer for Aspire Universal, the venture company that supports Tempo Health. Tempo is a mobile health start-up focused on developing new personalized technologies for diabetes patients.  Think of it as the most the sophisticated bolus calculator you ever saw, but one that does the math for you, taking into account not just carbs and insulin, but heart rate, skin temperature, altitude and more.

Personalized diabetes also means dynamic diabetes. An app I’m testing this week actually has four different blood glucose models running constantly, and constantly updates itself to use the most accurate one with each calibration.

Before you run from this post thinking that Kerri sold this space to the highest bidder, relax.  [Editor’s note:  The bids weren’t that high.  Second note:  This is a joke; no money was exchanged for this post, nor would there ever be.]  I’ve got nothing to sell you, yet. But I can give you my view of what changing diabetes looks like from this side of the blank white board and what’s required to take an idea and turn it into a useful product for diabetes treatment.

Tempo Health has a mandate from our VC to apply machine learning and the Internet of Things (IoT) to find new solutions for diabetes management. Simply put, we pull data off of devices (like a Dexcom G5 or a FitBit) and run it through seriously crazy math to predict your future BG. The most important word there is “your.” Every T1 knows that everything changes our BG every day. Some days our bolus calculations are spot-on, but often they’re not.

Assuming we can achieve personalized diabetes management with machine learning, there are many ways we could deploy the solutions: better bolus calculators for insulin pumps, non-invasive CGMs, etc. As the IoT continues to grow, so will the data and so will the insights that come out of the data.

When I first heard about using machine learning this way, I was taken aback. Months later, I can tell you that it’s not the math that’s holding back diabetes innovation. Here are three immediate hurdles that come to mind (relax, FDA – I’m giving you a pass this time around):

  1. Data. Everyone operating in the space needs data. Data from trials. Data from partners. Data from you. But truth is, we need more. Last I counted, there were twenty-four obvious factors that affect blood glucose, but the most robust integrated databases today usually include only three: carbs, insulin and BG. We need more data, which means…
  2. Willingness to share data. Vendors need to work together and share data with each other in joint ventures to help solve the problem. If you keep your chocolate bar and I keep my peanut butter, we’ll never have my favorite way to treat a low.
  3. Quantifying the potential value of personalized solutions is complex. If approaches to diabetes management are unique to each person, be it with data or devices, the value proposition per person is also unique.

The most exciting part about those obstacles is that while these are the things we bang our heads against the wall about every day right now, they don’t for a moment strike me as insurmountable. In fact, when I look at what’s going on in the industry, I see overcoming them as inevitable. Better technological approaches to diabetes treatment are coming soon and if you’re interested, you can be part of the revolution.

On our website, we have a simple form where you can sign up to be a beta user for our products. Like all studies, criteria vary. Sometimes we want people local to our Lancaster, PA offices and sometimes we don’t. Sometimes we’re trying out apps for pump patients and other times we’re looking at pen users. So I can’t promise we’ll use you, but we’d love to have you sign up today. Just go to http://tempohealth.com

Marcus Grimm is CMO for Aspire Universal, serving as a startup storyteller for multiple ventures. He’s also the Running Coach for Diabetes Training Camp is on the Board of the Diabetes Sports Project and his local JDRF. Follow him on Twitter at @marcusgrimm.

Share: