Inspiration
Diabetes is a major public health concern worldwide that is projected to increase to 592 million by 2035. Approximately 8–18% of diabetic patients with cancer experience poorer prognosis, survivorship, and quality of life. The long-term survival likelihood for diabetic patients with cancer are 1.4 times less compared to cancer patients without DM. Managing diabetic patients with cancer is challenging because insulin therapy puts them at high risk for hypoglycemic events that can be life-threatening, some types of chemotherapy may impact blood sugar levels, and patients receiving cancer treatment often require pain management. Therefore, diabetic patients with cancer require close monitoring and timely access to care to maximize treatment outcomes and improve their quality of life. Currently, there are no wearable products that exist to support timely management for diabetic patients with cancer.
What is G2 Glove-Pump
To address this problem, our project will tackle blood glucose levels and pain management in diabetic patients with cancer. Modeling off the current system of the insulin pump and continuous glucose monitoring (CGM), our product will combine the use of a quick detection system and response algorithm for both blood sugar levels and pain medication administration. Our detection system will utilize blood glucose levels as the biomarker to measure changes in patients glycemic index and the Galvanic Skin Response (GSR) sensor will detect changes in pain perception. Alerts would be incorporated into the product signaling when values exceed a critical threshold that require immediate medical attention. Other data such as pulse, oxygen saturation (SpO2), body temperature, and insulin/pain reliever dosages will also be collected. Data will be available both on the physician’s side as well as the patient’s. With the accumulation of this data, clinicians can be able to see which patients are doing better on which medications and determine appropriate dosing adjustments, depending on what trends they see. This can become an asset used to help build on precision medicine and shared-decision making.
How we built it
| Picture | Explanation |
|---|---|
 |
For hardware, we try to prototype it using 3 of these sensors: Pulse SPO2, Body Temperature and GSR controlled by Arduino Uno. |
.png?alt=media&token=5db8d91d-b383-4b8f-ae6a-e0bb6dc51f4e) |
For software, we used a react framework called Next.js served as a PWA to build our frontend. The Arduino wearable technology transmits live data to our MongoDB Atlas Database and we subscribe to the live data on our frontend. We show real-time data of Glucose Levels and GSR Deviations in the dashboard. |
Integration of Software & Hardware
The system draws data from an array of hardware sensors and the DEXCOM API provided by the Dexcom cloud. We use the Dexcom API to provide us with sandboxed values of blood glucose levels (as we currently do not have any team members with an attached continuous glucose monitor). The remaining hardware consists of a MAX30102 sensor which uses LEDs to detect pulse and SPO2 (blood oxygen saturation) levels; a Galvanic Skin Response sensor with attached ADC to detect galvanic skin response/impedance and an analog body temperature probe for observing the body temperature of the patient. The Max30102 is similar to COTS pulse oximeters available and used in hospitals. These sensors are connected to an Arduino Microcontroller board which calibrates and collects the sensor readings, then uploads them to a Mongodb database via serial communication. In the next version, we will replace the board with a smaller and more powerful board like the ESP32 which can communicate via Wifi and Bluetooth for a more portable prototype which can be worn like a glove. The Mongodb has triggers which can call a remote function that alerts caregivers/physicians if any of the observed parameters move out of a safe range for a period of time (> 4 observation window)
Challenges we ran into
A challenge we ran into was pinpointing which population we we wanted to target and how to make it a desirable product in the market. We overcame this by conducting literature review and curated our story using literature evidence and signifying the importance of our glove-pump system. We also ran into a challenge when creating an efficient algorithm to check the previous occurrences of both GSR Deviations and Glucose Levels and contact the physician if there are any abnormalities.
Accomplishments that we're proud of
We are proud of designing software and hardware that sync up together to detect specific values and proposing the plan of using these values to reflect an appropriate action.
We are also extremely proud to create a solution that is 63.08% cheaper than the existing solutions on the market. Not only did we build a solution that is affordable, but we also have the cheapest solution on the market by a large margin.
What we learned
Diabetic patients with cancer have poorer quality of life, prognosis, and survivorship compared to diabetic patients without cancer. They require close monitoring and timely intervention. There are limited person-centered solutions to support patient self-management of the side effects of these conditions.
Research
Paola Ballotari, Massimo Vicentini, Valeria Manicardi, Marco Gallo, Sofia Chiatamone Ranieri, Marina Greci, Paolo Giorgi Rossi: Diabetes and risk of cancer incidence: results from a population-based cohort study in northern Italy, 2017 Oct-Dec, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5657107/
Denise Soltow Hershey: Importance of Glycemic Control in Cancer Patients with Diabetes: Treatment through End of Life, 2017 Oct-Dec, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5559941/#:~:text=Cancer%20patients%20with%20diabetes%20are,studies%20have%20explored%20this%20hypothesis.
Log in or sign up for Devpost to join the conversation.