Backstory

Before attending college, I had a routine physical to make sure I was healthy enough to participate in club sports. I'd been very active throughout high school (captain for both the lacrosse and swim teams), so I wasn't expecting anything to come up.

Still, the blood tests came back with high blood sugar levels. As it goes with most diabetics, I was retested, had similar results, and was eventually diagnosed with type 2 diabetes. Diabetes is passed down genetically in my family, so the diagnosis wasn't all that surprising, but I still struggled with accepting my diagnosis and changing my lifestyle to better manage the disease.

Four years later and now I'm a senior computer science major who regularly competes in hackathons. With my background in technology and first hand experience of managing type 2 diabetes, I'm excited to participate in the HITLAB World Cup.

Problem Statement

Successfully managing diabetes involves exercising, counting carbs, regularly self-testing blood sugar levels, taking medication... the list goes on. On the other hand, there are only two available metrics that measure how well a person is managing their diabetes: daily blood sugar readings and the periodic A1C blood test. The daily tests help analyze blood sugar levels but are heavily dependent on factors like meals and exercise, so if used incorrectly, they can provide a false sense of assurance to diabetics. A1C test are useful because they basically measure a 3 month blood sugar average, which avoids the fluctuations based on meals and exercise, but if only taken every 3 months, then patients are left in the dark for long periods of time.

Hello Diabetes will attempt to solve this problem by using voice recognition technology to aggregate information about a user's food choices, exercise habits and blood sugar levels to ultimately make an estimation of the user's current A1C. By providing an estimated A1C score, diabetics won't have to wait 3 months before receiving feedback on how well they are managing their diabetes. The application will also provide a user with insight on what is impacting the A1C estimation so they can improve their behavior before the real A1C test results. This application is not meant to replace the A1C test; instead, it is meant to supplement the test and help diabetics attain better scores.

The application will use voice technology to allow the user to enter all relevant information hands free. The user could simply tell the app "Hello Diabetes, my current blood sugar level is 125". The app could then respond, "Great, was that reading taken before or after a meal?"; thus prompting the user for additional information. As information is gathered and a user profile is built, the user could begin to ask, "Hello Diabetes, what does my current A1C estimate look like?"; to which the app would respond, "Hi Bill, your current estimated A1c is between 6.5 and 7.5 based on your recent decrease in average blood sugars and improved eating habits. Would you like to hear more about the estimate?"

Impact

Recording blood sugar results, meals, and exercise is time intensive and boring. It may also feel meaningless to a diabetic especially when they don't understand the full impact of their decisions. By using voice technology, the user can log the values without opening a notebook or phone application and receive instant, interactive feedback on their decisions. This instant feedback and will empower diabetics to make better decisions before doctor's visits, not just after. This app is primarily targeted towards the diabetics themselves, but there is no reason why a 3rd party caregiver couldn't use the application as well.

Innovation

Many applications are available today that log blood sugars, provide nutritional information, and monitor exercise. This app distinguishes itself by...

  1. Estimating A1C values based on user inputs
  2. Detailing how the A1C estimation is impacted based on recent inputs
  3. Providing a convenient hands free interface... no pencil or phone needed to interact with application

Sustainability

In order to pay for the cost of maintenance and improvements, the software's user would have to pay a small annual or monthly subscription fee. The importance of educating diabetics and pre-diabetics about how to manage their disease better can't be underestimated. I think if this app is pitched correctly, it could definitely elicit funding from government agencies and healthcare providers.

Feasibility

This app would leverage pre-built voice technology libraries, which are relatively easy to use. The hard part would be creating an accurate A1C estimator that provides valuable insights to users about how to manage their diabetes. While I have first-hand experience with managing diabetes, I would definitely need to work with diabetic professionals to create a valuable user experience. While I'm very self-motivated, I'm still always willing to ask for help. As this application develops, I'd be ready to add members to my team who can provide the technical skills and medical expertise to expand the app. Partnering with larger pharmaceutical corporations for mentor ship and funding is also a possibility.

Share this project:
×

Updates