A physician-turned-mobile app developer: My first hackathon

Brian Rosman, MD, is the current Robotic Surgery Research Fellow in the Department of Urology at Boston Children’s Hospital. He focuses on applying modern technology to improve the practice of medicine.

When I entered the MIT Health and Wellness Innovation “hackathon,” it wasn’t with high hopes. I felt our team from Boston Children’s Hospital had a solid idea to develop, but I’d taken part in similar collaborations that fell victim to personal disputes, group member apathy and overzealous leadership. This time I was wrong.

The idea we brought to the 10-day event/competition, sponsored by MIT Media Lab, was a mobile telecommunication robot that could monitor children in their homes after surgery. Our mobile robot, the VGo, has the potential to be a great tool, but when we’re not using it to make remote visits, it just takes up space in the patient’s house.

We felt the robot could be enhanced if the family could interact with it to ask questions, through some type of artificial intelligence, while the child played with some kind of game or character that would help them stay on their post-operative regimen.  The machine already had digital eyes, ears and legs. We wanted to give it a brain.

That became the idea that we submitted to the hackathon. As clinical lead on the project, my role was to bring clinical context to our idea and create a model patient. Elizabeth Phillips of the Innovation Acceleration Program at Children’s came along to help lead the project, along with two wonderful programmers from the FastTrack Innovation in Technology program, Gajen Sunthara and Melinda Tang.

Day 1: Me, pitching our project: a mobile robot that monitors children in their homes after surgery and motivates them to follow their post-op regimens. (Photo: Paul Franzosa)

Once we arrived, several volunteers signed on to help us out, including David Furnivall of the National Health Service Trust in London; Uri Feldman, a Spanish interpreter for Children’s Hospital Boston and a graduate of the MIT Media Lab; and Paul Franzosa, an independent engineer/entrepreneur.

These people came from many backgrounds, and brought many different kinds of expertise, which they quickly matched to different parts of our project. Everyone contributed and respected one another in a constructive work environment. There was no need for a dictatorship or endless deliberations, because everyone was united around the same goal. My cynicism evaporated immediately.

As a physician, I work and think in very concrete terms, but as we got started, our concept was still nebulous. The diversity of experience in our group was essential in turning it into a whiteboard full of screenshots and branching algorithms that would lead to a tangible product.

Day 1: The hackers get to work. (Photo: Paul Franzosa)

Our project took the form of a web application, but our group had only two programmers who could turn our ideas and drawings into actual working code.  The challenge became targeting the efforts of the non-programmers to accomplish as much planning and designing up front as possible, to make things as simple as possible for the coders.

Some group members people created screenshots of some of the parts of the program that we didn’t have time to code (I was shocked at how little time 10 days is); others focused on how to organize our various ideas into an easy-to-use interface.

One question we discussed was how best to incorporate an avatar that would engage and motivate children to take their medications and record important metrics at the right time. There were several excellent ideas, from a Tamagotchi style character that the child would have to feed and play with to keep it healthy, to a board game approach where the child would move the character around the board as medications were taken.  We finally decided to let the child customize his or her avatar’s features and accessories, making the ability to buy new upgrades contingent on compliance with the post-operative regimen – similar to the way Club Penguin lets kids buy decorations for their penguin’s igloo, but only if they become members.

A rough prototype of our app, a mobile robot that monitors children at home after surgery and motivates them to follow their post-op regimens. The clock face displays a schedule for medications and documentation to fill out throughout the day. (Photo: Paul Franzosa)

There were even cautions against having the child get too invested in the game, lest they be devastated when the game is taken away after the post-operative period, or swallow an entire bottle of pills in an attempt to get more points and accessories!

At the end of this collaborative process, the non-coders had a well-defined idea that the coders could run with.

I have heard people talk about the satisfaction of developing an idea before, but my first experience with it, as our idea came to life on the screen, was overwhelming. I felt an enduring sense of pride in our group and our idea.

Paola Abello, program manager for the Innovation Acceleration Program, poses as a frazzled parent for our video demo. (Photo: Lindsey Hoshaw)

After many long days of whiteboards and coding, the competition day arrived: We had to create a 90-second video, and then demonstrate our program during several hours of exhibition for judges, our competitors and industry representatives.

It was surprisingly difficult to compress eight days of work and a tremendous scope of the project into 90 seconds. Boston Children’s sent over some reinforcements, and we pressed them into work as actors. Working with the video crew was both enjoyable and stressful: I loved how they were able to showcase the best parts of our prototype, while minimizing the parts that hadn’t been polished, but I was also reminded of all of the compromises we’d had to make.

After the videotaping, we set up our display table and tried to figure out the best way to showcase our program. We had tablet computers to show it on, but our original idea was for the application to run on our mobile telecommunication robot. Since the robot isn’t currently compatible with the software, we ended up simply using Velcro to affix the tablet to the robot’s front, allowing us to drive it around the room and show off the application in its envisioned use. It was a tremendous hit and drew people’s attention to our table and our program.

Frank Moss, head of the New Media Medicine group at the MIT Media Lab, talks to me through the VGo robot, the same way a patient would. (Photo: Lindsey Hoshaw)

At the end of the day, we were thoroughly sick of talking about our post-operative care application, but very satisfied with the results of our efforts – especially when we were announced as winning third place. We were elated at having created a product that was recognized by industry leaders as useful and innovative.

The exhibit room on the final day of the hackathon, showing the six teams' demos (Photo: Paul Franzosa)

This experience taught me that a dedicated group of people, all striving towards the same goal can accomplish great things.  I had never met any of the members of the group before that first day, and none of us individually had all the skills to meet the challenge. But together we pulled together a concept out of thin air and created a viable and exciting piece of health care software – in just 10 days.