STEM-In-Action Fall Follow Up: Stumptown Glucobots

As another week goes by, another STEM-In-Action team is amazing us here at Mission Control. We look forward to our weekly updates on each of the five grant recipient teams, to see how they are progressing with their projects and helping their communities. If you are new to eCYBERMISSION or don’t know about our STEM-In-Action grant, the U.S. Army Educational Outreach Program (AEOP) STEM-In-Action Grant awards eCYBERMISSION teams up to $5,000 to develop their projects into mature and scalable solutions in their community. We have already caught up with two of our five teams, and are now heading to the west coast to catch up with the Stumptown Glucobots!

--

We are the team “Stumptown Glucobots”, a neighborhood-based team from Portland, Oregon, comprised on 7^th graders attending Stoller Middle School. We are on a mission to build a solution to measure blood sugar levels without a drop of blood!

According to the Center for Disease Control and Prevention, nearly 100 million U.S. adults are now living with diabetes or prediabetes, and that number is expected to grow substantially by 2025. The amount of children with juvenile diabetes (type one) has also grown to a staggering 200,000 kids, which is a dramatic increase compared to past years. Diabetes is an illness that occurs when insulin, an essential hormone for providing glucose, (the life energy to cells) is not produced in sufficient amounts by the pancreas. This usually results in fainting, vomiting, fatigue, and even death. Presently, the only FDA approved way to measure blood glucose is to invasively prick your finger, which involves a syringe extracting blood from an area of designated skin, usually the fingertip, and coating this blood sample on a thin enzyme-coated strip and then inserting the strip into a glucose monitor. However, the large population of children finds themselves irritated and sore, and according to medical professionals, this results in very big tantrums and disruption in the clinic or hospital. This method using strips and syringe for blood extraction is extremely expensive and costs over $700 yearly.

Using the eCYBERMISSION website and other resources, we came up with this Problem Statement: What can be done to cost-effectively and reliably prevent diabetic children from having to prick their finger daily?

We subsequently started researching with help from our relatives and friends in the medical field and set out to create a cost-effective, non-invasive blood glucose monitor without a drop of blood. Our team collected useful information from more than fourteen different sources ranging from books, websites, case studies, articles, and expert discussions. We also met with a number of doctors, subject matter experts, a researcher, and diabetic patients in various age groups to analyze the importance of this global problem, including those from Providence Health Group and Oregon Health and Science University.

We then started collecting more than fifty data samples from diabetic patients, multiple times. We spoke to Mr. Jaspreet Jhoja, a Masters student at the University of British Columbia to learn about using a Near Infrared (NIR) sensor to measure the glucose value. We analyzed various nanometer spectrum level and conducted more than six different tests to identify 860 nm wavelength based Near Infrared sensor from Sparkfun.com and built a prototype to demonstrate a non-invasive solution for measuring blood glucose levels. Using more than fifty data samples we created from real patients, we created an Artificial Intelligence-based Linear Regression model to correlate measured glucose values with corresponding NIR sensor values, and predicted blood glucose levels with more than 85% accuracy, without a drop of blood, using our solution, Glucobox.

By collecting more data samples from various patients, we can improve our machine learning model and the prediction accuracy to more than 95%. With the STEM-IN-ACTION grant, we plan to build additional Glucobox devices using 3D printing technology to setup Glucobox at various hospitals in our neighborhood to collect more samples, and subsequently, shrink the form factor to a device that can be attached to Android phones.

Our solution benefits the community by allowing millions of the diabetic population to create and use a painless solution that can quickly be used. Even though our initial focus was on juvenile diabetes, through our research we have proven that our solution, Glucobox, has the potential to benefit all diabetic and pre-diabetic patients worldwide. This is a vast improvement from the painful and time-consuming invasive monitors that are used throughout the world. We are planning on posting the programming and parts required onto the internet so anyone who wants to buy, assemble and program the parts, if they so choose. Our solution is under $200 which makes it more affordable. We will work to refine our model to find the most efficient parts after the first round is completed so we can have additional time to test different components of our device. In addition, since our solution is non-invasive and does not require a blood sample, we can prevent any spreading of contagious diseases in our communities.

-Team Stumptown Glucobots, Portland, Oregon-

Shreyas Ananth, Kapil Kakodkar, Rishab Madhusudhan

--

Again, a team that set out to help their community, but have a reach that has the potential to have a global impact. We are so excited to see this team’s progress and cannot wait to catch up with them in a few months.

-Mission Control