About a year ago I got an email from a colleague in radiology about a project he was beginning that was designed to trained midwives in Uganda to use ultrasound to try to reduce maternal mortality. The premise of his project is that there are three conditions that contribute to a significant portion of maternal mortality (through hemorrhage and infection) which could be relatively reliably identified using ultrasound.
Nearly a year later, I can say that partnering with him has been the best thing I’ve worked on in recent memory. While his original project focused on training midwives in basic ultrasound and then providing them with commercial portable ultrasound units, at some point it became clear that a new kind of technology — something that was cheaper and easier to use — would potentially make a considerable difference.
Since we were doing this with no funding, I took it to my HCI class as a project topic. What happened was kind of magic. It’s been about ten months since this started as an undergraduate project. We have a working prototype that builds on a commercial USB probe put out by a company called Interson and an SDK developed by researchers at Washington University. We’ve simplified the user interface radically, taking out some of the functionality of ultrasound that experts perceive as core, but which aren’t actually essential to the level of diagnostic work we need done by the midwives for the goals of maternal health. And the students had the idea of a robust, contextual help system. Since the midwives generally practice in clinic settings where they are on their own, there is no one to consult when they have a question about images or diagnoses. Cell connectivity is close to nil, so that leaves them isolated with no community of practice on which to draw.
There are two things I love about this project. First, the students are amazing and inspiring. Second, though, it’s a kind of validation of my assertion that non-experts are capable of great innovation and creativity, and that entrenched problems can potentially be solved by people well outside any traditional notion of expertise. Lots of people have worked on the problem of low-cost ultrasound. It took a group of undergraduates, though, to radically rethink some basic assumptions and be willing to articulate an unconventional approach.
Also, I’ve been ranting for a while about the virtues of older and simpler technologies, and that’s exactly the ethos of this system.