The goal of this website is to allow a user to track the carbs they ate during a meal, along with the fluctations of their blood sugars, tracked with Dexcom. Dexcom is a continous glucose monitor that attaches to a diabetic's arm, and reads your glucose levels, every 5 minutes. The Dexcom API makes the glucose readings, calibration events, data range and user events available to third party apps.
While a Type 1 Diabetic calculates the number of carbs they eat in every meal, so they can properly dose their insulin, a diabetic's blood sugar may still end up going too high or too low. I wanted to allow a user of this app to see carb counts for a meal along with a line graph of their blood sugar fluctuations for that day. I’m using the Edamam API to allow the user to find the carb count of the foods they’d like to eat, and to upload the data for each meal (similarly to My Fitness Pal, but with a focus on carbs). I’m displaying the carb counts for a meal along with a line graph of their blood sugar fluctuations for that day. I’m hoping it will give the user visual insight into how their meal affected their glucose readings. I plan to continue by studying machine learning. I’d like to train the AI with the specific foods a user eats, how it affects their blood sugar, and then use that insight to time insulin injections. I think machine learning would be quite effective at helping someone avoid highs and lows.
I made a live demo version of the app, using mock data from Dexcom, so that the site can be used by someone who doesn't have a Dexcom account Demo with sandbox data: https://dexcom-tracker.herokuapp.com
Technologies used: