Every day, I look at my insulin pump screen and push the buttons to give myself the medication that keeps me alive. I’m lucky, because I can see what I’m doing. But many can’t. And with diabetes, I am afraid that someday, I might become one of those folks who can’t see well enough to use this life-sustaining device that I’ve depended on for 14+ years now.
Obviously, with diabetes being a major cause of creating vision problems, many of us would be well-served by talking D-devices — now and in the future.
That’s why it’s exciting to see a project at Michigan State University in which a small team of student researchers are creating a voice chip that can be installed in an insulin pump. At this point, the existing prototype uses the Asante Snap pump, thanks to that California company being the main backer of this research.
Four college seniors at MSU’s Department of Electrical and Computer Engineering spent the past several months developing this solution for the visually-impaired with diabetes. As of now, the voice chip simply reads off whatever is displayed on the pump screen, but in the future they hope to have full audio features incorporated. In December, the students finished and published a 57-page report on the project.
If all goes as planned, the MSU student-led team hope to publish their research in the Journal of Diabetes Science and Technology to help other engineers design pumps or other medical devices for people with vision disabilities.
“Their proof of concept helped uncover and address many design issues that one will face in making pumping accessible to the visually impaired,” says Mark Estes, chief product architect at Asante. “We love the project, were glad to support it, are in discussion with them on how to keep the ball rolling, and most importantly we are very impressed by the team’s output.”
A Call to Action
Surprisingly, the impetus for this project didn’t come from any PWD (person with diabetes), as there aren’t any on the research team. The man leading the charge is Stephen R Blosser, assistant technology specialist at MSU’s Resource Center for Persons with Disabilities. During his college days, he used to volunteer with the state’s Bureau of Services for Blind Persons and has kept in touch with those advocates ever since, so when they reached out about the need for more accessible devices for people with diabetes, Blosser was quick to take action.
He doesn’t have any personal connection to diabetes, but the stats were hard to argue with, he says — approximately 40% of PWDs in the U.S. have some form of retinopathy, but as of now the insulin pump industry isn’t addressing that group.
He gathered four interested students for this project, and they reached out to well-known CDE Gary Scheiner in Philadelphia, who helped connect them with Asante. Here’s how Scheiner tells us he reacted:
“I loved the concept. That’s what society is all about: people with special skills/needs doing what they can to help others who have their own special skills/needs. That way we all benefit from the best of the best. I put the team in touch with Mark Estes at Asante Solutions, makers of the Snap insulin pump. I’ve had an opportunity to work with the developers for virtually every insulin pump manufacturer in some capacity, and Asante really stands out to me as a company that is adaptive and flexible to consumer needs/interests. Despite the fact that a project like this probably won’t help the company’s bottom line, Asante is working with the student team and making tremendous progress. We’ve never had an insulin pump that meets the needs of people with limited vision, and there is a significant number of people within the diabetes community who have visual impairment and would benefit from pump therapy.”
Blosser tells us he also reached out to Medtronic about doing the same with their Minimed insulin pumps, but never received a response from the company.
Asante agreed to participate and donated several insulin pumps and batteries for the research, as well as other tools and tech support to help the students understand what was going on inside the pump’s controller.
“The students did the vast majority of the work,” Estes tells us. “If there was a pie chart of who did what, you would not be able to see our slice. So, they deserve huge props for their work.”
The Talking Pump
At the start, the researchers rejected both the ideas of adding Braille to the buttons and also using Bluetooth to connect the pumps to smartphones for the talking component — in part because some may not be able to fully feel those buttons thanks to loss of fingertip sensation or lack of knowledge of Braille, and also because of security concerns in sharing patient data via smartphone. Instead, they opted to alter the buttons by connecting a touchpad sensor to a speech chip inside, triggering the speech when the button is pushed. So far, the group has come up with two possible configurations for the pump hardware — one that has the voice feedback built in, and a second that would have an add-on speaker module.
By design, the pump would vocalize any information — from the button you’re pushing, to the dosing and carb calculations you enter into the device. Blosser says some of the early research was limited because they didn’t modify the pump software, as they were only working to make the device accessible. Going forward, the team will expand that R&D to address the most “universally usable method for operating an insulin pump” so that it can vocalize any information needed on the device, and could be easily adopted by any pump manufacturer.
You can view a quick video demo of the talking pump going through some of the basic button-push menu screens, with the audio being one of the research students who recorded her own voice saying all of the commands and functions to give it that “human element.”
Blosser tells us they hope to finish an operating virtual pump by the end of this Spring and then to conduct usability tests with blind or visually-impaired users this Summer. But the timeline depends on funding, he says, and the cost could be about $20,000.
“It would be my first priority to make this open source for any pump manufacturer or even any general medical device designer to adopt. However, Asante has sponsored the project thus far and we have signed nondisclosure agreements (for their pump hardware). They should be given first choice of adopting the design, and the option to patent any novel ideas,” Blosser says. “I feel that most of what we will do will not be novel but public knowledge and not patentable. But this doesn’t mean it’s not valuable. Our usability and research laboratory will probably be the place where the most valuable discoveries will be made.”
As to how this fits into Asante’s pipeline, Estes tells us: “While this project is laying the foundation for what we hope is broader access for all users, there are no immediate plans in terms of implementation into production that I can share.”
Addressing a Blind Spot
We’d bet that many D-Community living with visual impairments would be quite excited at the possibility of access to a talking pump. In fact, one of our past DiabetesMine Design Challenge winners who is legally blind did a video a few years ago titled “Do Diabetes Technology Companies Have a Blind Spot?”
We also chatted with DOC friend Matthew Deets in Oregon, who lives with T2D and is on multiple daily injections — and happens to be legally blind, being born with Optic Nerve Hypoplasia (ONH) in both eyes.
“I’ve always wanted to try pumping, but it’s been out of the question for me not only because of the accessibility issues, but because my insurance company really basically doesn’t cover pumping or CGM,” Matthew says. “I have always thought it’s strange that insulin pump companies have not included optional voice adaptation for accessibility in their equipment because unfortunately complications relating to diabetes is one of the leading causes of blindness worldwide. I witnessed this firsthand with a relative of mine who has been an insulin pump user for years, but is losing his eyesight and will most likely have to switch to MDI. On most pumps, there’s really no way for a visually impaired or totally blind individual to make their own adjustments.”
The idea of this talking insulin pump is something Matthew says is long overdue and a needed tool, and he’d be interested in helping out as much as possible.
Blosser says they’re very interested in hearing from anyone in the D-Community who has thoughts or insights to share on this project, or even wants to help trial-test some of the developments. He says: “This project is a chance for anyone who uses a pump to make suggestions for improving its operation.”
For our part, we definitely hope that Asante chooses to take on this incredible research for its own device design, and that other device makers are able to follow suit.