MIT’s all-in-one approach to treating diabetes – Includes an app that identifies and quantifies food content

MIT engineers are working on an app that identifies and quantifies food content, which can help count carbohydrates for people with diabetes. Credit: MIT News, archive footage

MIT Engineers are working on a new type of device that could streamline the process of measuring blood sugar and injecting insulin.

Before consuming a meal, many people with diabetes need to inject themselves with insulin. This is a time-consuming process that often involves estimating the carbohydrate content of the meal, drawing blood to measure blood sugar, then calculating and administering the correct dose of insulin.

These steps, which usually need to be repeated with every meal, make it difficult for many diabetes patients to stick to their treatment regimen. A team of MIT researchers has now developed a new approach to streamline the process and help patients maintain healthy blood sugar levels.

“Any intervention that makes it easier for patients to receive therapy can have a huge impact, as there are multiple barriers related to time, inconvenience, dexterity or learning and training,” says Giovanni Traverso, Karl van Tassel Career Development. Assistant professor of mechanical engineering at MIT and gastroenterologist at Brigham and Women’s Hospital. “If we are able to overcome these barriers through the implementation of new engineering solutions, it will be easier for patients to receive this therapy.”

Traverso and his colleagues have designed two different devices that can simplify the process of calculating and injecting the right dose of insulin. One, which combines several of the existing steps into a single device, could be used in patients in the near future. Their second prototype incorporates flexible electronics on the surface of a needle so that blood measurement and insulin delivery can be done through the same needle. This could possibly make the diabetes management process even more streamlined.

MIT postdoctoral fellows Hen-Wei Huang and Sean You, and visiting students Luca Di Tizio and Canchen Li, are lead authors of the paper, which appears in the Controlled Release Diary.

All-in-one device

Diabetes affects 34 million people in the United States and more than 400 million people worldwide. Patients with diabetes often use two types of insulin to control their blood sugar: long-acting insulin, which helps control blood sugar over a 24-hour period, and short-acting insulin, which is injected at mealtimes . Patients first measure their blood sugar using a blood glucose meter, which requires pricking their finger to draw blood and placing a drop of blood on a test strip. They must also estimate the amount of carbohydrates in their meal and combine this information with their blood sugar to calculate and inject the correct dose of insulin.

Existing technologies such as continuous blood glucose monitors and insulin pumps can help with parts of this process. However, these devices are not widely available, so most patients must rely on finger pricks and syringes.

“Every day, many patients have to perform this complicated procedure at least three times,” says Huang. “The main objective of this project is to try to facilitate all these complex procedures and also to eliminate the need for multiple devices. We also used a smartphone camera and deep learning to create an app that identifies and quantifies food content, which can help with carbohydrate counting. »

The research team designed two different types of “all-in-one” devices, both of which incorporate the new smartphone app. Patients first take a picture of the food, and the app can then estimate the volume of food as well as the amount of carbs, based on nutritional information from a USDA database.

The first all-in-one device the researchers designed consolidates many existing tools that patients currently use, including a lancet to draw blood and glucose test strips. Once the blood glucose measurement has been taken, the device transmits the information to the smartphone application via Bluetooth, and the application calculates the correct insulin dose. The device also includes a needle that injects the correct amount of insulin.

“What our device does is automate the procedures for skin pricking, blood collection, glucose level calculation, insulin calculation and injection,” Huang explains. “The patient no longer needs a separate lancing device, blood glucose meter and insulin pen.”

Many components included in this device are already FDA approved, but the device has not yet been tested on human patients. Tests on pigs showed that the system could accurately measure glucose levels and deliver insulin.

One shot

For their second device, the researchers wanted to come up with a system that would only require one needle stick. To achieve this, they designed a new glucose sensor that could be integrated into the same needle used to inject insulin.

“The idea would be that if we could integrate the glucose sensor directly onto the surface of the insulin delivery needle, we would only need one rod for the patient, which minimizes pain and also makes it easier to administer the whole process,” says You.

The researchers designed a flexible electronic sensor that can be attached to the needle and measure glucose levels in the interstitial fluid, just below the surface of the skin. Once the needle enters the skin, it takes between five and 10 seconds to measure glucose levels. This information is transmitted to the smartphone application, which calculates the insulin dose and administers it through the inserted needle.

In tests on pigs, the researchers showed that they could accurately measure glucose levels with this system and that glucose levels dropped after insulin injection.

Because this device uses a new type of glucose sensor, the researchers expect it will require further development to get to a point where it could be tested in patients. They have filed patents on the two systems described in the new study and hope to work with companies to develop them further.

The research was funded by MIT’s Department of Mechanical Engineering and Brigham and Women’s Hospital.

About Stuart M. McFarland

Check Also

Indiana schools use app to connect students in need to community donations

Low-income students may struggle to meet basic needs, such as new clothes and shoes. Without …