Designing Patana AI: A New App for the Early Detection of Parkinson’s Disease
More than 10 million people are living with Parkinson’s Disease worldwide, according to the Parkinson’s Foundation. Early detection of the disease, which can cause tremors and make it very difficult to walk or move, can mean access to medications that can dramatically slow the progression of the disease. But long before most people will ever talk to their doctor about the possibility of having Parkinson’s Disease, they may show symptoms; changes in their posture, gait, and movement. Catching these changes is crucial to getting an early diagnosis.
For Babusi Nyoni, founder of design firm Tripleblack Agency, the opportunity to use technology to help with this early diagnosis was apparent. In 2018, he and his agency developed an app for the South African market, using AI and computer vision, to help users rate a particular dance move. What started as a project to create a fun app and test South Africans’ willingness to interact with new technologies morphed into a much greater cause.
“We started exploring ways to use the formula we made in a more impactful way. And we, after doing some research, discovered that Parkinson’s Disease diagnosis relies partially on symptomatic observation over time of your posture and your gait and your movement, which we had already learned how to parameterize with the dance app,” said Nyoni.
Over the past year and a half, Nyoni and his team, spread between Amsterdam and Zimbabwe, focused on repurposing their technology to aid in the early diagnosis of Parkinson’s. Read on to learn how they developed and tested Patana AI, now launched, with the community it serves, using Adobe XD along the way to design and prototype the experience.
From user testing to “co-innovating” with impoverished communities
Tripleblack Agency aims to use emerging technologies to solve problems in the Global South. By designing for impoverished communities first, Nyoni says the agency is able to guarantee their products will be useful and usable for everyone, regardless of level of privilege. Working out of his office in Amsterdam, he collaborated with team members on the ground in Zimbabwe as they observed, tested, and tweaked machine learning models that could look at subjects’ posture, gait, and the presence of tremors. To do this, they worked closely with health care workers and residents at long-term care facilities for seniors.
“What we’re doing, essentially, we’re taking this part of the assessment that relies on human perception and we’re quantifying it, and we’re also trying to make the experience one that adds as much value as possible to a doctor,” said Nyoni. Patana AI doesn’t aim to replace a doctor in the diagnosis process, but to provide a tool to help encourage people to seek help and to aid the doctor with their diagnosis.
Throughout Tripleblack Agency’s extensive user testing, they uncovered many insights into the needs and limitations of seniors that fed back into iterations of the product. Reducing the number of steps a user needs to take to use the tool to successfully assess their gait, for example, was important to avoid falls. During this process, Nyoni realized he needed to think of his “users” more as “patients.”
“Think of the thickness of the syringe, for example. You could get as much liquid in as possible with a thicker needle, but then that wouldn’t be the most comfortable experience for a patient,” he said.
Then, there’s the question of the results of the test. Tripleblack Agency spent a great deal of time working with seniors, to make sure the end results (assessment of whether the user is showing symptoms of Parkinson’s) were communicated in a way that was clear, and gave them exactly the information they were looking for. Getting the end results part of the experience wrong would destroy the whole user experience for its key users. This required a lot of trial and error, and iteration, in XD.
“Having people who worked in a nursing home in a small town in Zimbabwe, taking part in the formulation of a solution that’s first in its field and built solely for them…this is an important takeaway. We co-innovated with people from an impoverished background, on something that is potentially so important for the rest of the world, and made it a focus on them first. This is the key to designing for the whole world,” Nyoni said.
Key considerations when designing for the Global South
Effectively designing for seniors and their caregivers, especially regarding health, was a process that already came with unique UX constraints. These potential users, “patients” as Nyoni now prefers to think of them as, are already vulnerable, even in privileged circumstances. Tripleblack Agency, with its mission to design products with the Global South in mind, set out to ensure Patana AI would be usable in non-privileged settings, as well, such as impoverished communities in Africa. This came with its own set of user testing challenges.
“In some of our user testing sessions, there was no electricity in the home, which is a typical thing in Zimbabwe (due to blackouts). We had to work around that, using XD to communicate design decisions. It also made us realize we have to build an experience that, for example, would give the test-giver notifications on whether or not they need to improve the lighting conditions,” Nyoni said.
These “rolling insights” became commonplace. Many communities in Africa are just coming online for the first time, and they’re doing so because they now have access to mobile devices. For some users, even younger ones, they may face a cognitive burden in trying to use experiences, that those who grew up with the internet do not. This means designers need to approach “ease-of-use” in a new way.
Looking at Zimbabwe, specifically, Tripleblack Agency realized it had to create an experience that was both light on both power and data. Africa has the fastest growing adoption of the internet in the world, but it also has the most expensive data costs. Nyoni and his team considered building two versions of the app; a full-on experience and a data-light option. In the end, they focused on building one experience that would work in many contexts.
“With data costs, you need to be either a very high-value interaction or one that won’t use much data in order for people to come online. Sure, I would rather have an experience where you open the app, you complete the onboarding scheme, and then you’re presented with a camera interface and you can initiate the scan process. But just launching the camera automatically for someone already uses up battery and they might not have enough data to complete the experience anyway,” said Nyoni.
“So, we realized we really had to really stagger the process and make every single usage of data and battery as consensual as possible. These are very delicate things when you’re building experiences for people who have scarcity on both fronts.”
Getting buy-in from the medical community
There is a particular level of scrutiny, in general, for health and medical-focused apps and experiences. Patana AI is a tool to help early diagnosis, and assist doctors with this process – that focus on assisting, rather than replacing, has meant a level of openness and acceptance from the medical community. Nyoni says there is an excitement about new technology among doctors.
“I think we’ve had this really convenient cushion, which is the technology. There’s a general excitement around how this thing is going to work…If we created a new kind of paper-based notebook that helped tabulate these things, for example, I don’t think we’d have gotten the same reaction.” he said.
Designing for the vulnerable first, to accommodate all
“We’re trying to solve for the poor, and then the universe later.” This is a guiding principle of Tripleblack Agency, said Nyoni. For Patana AI, the experience has been optimized for the best usage in places like rural Zimbabwe, built from the ground-up with the challenges people face there in mind. It’s a challenge that really pushes UX designers: optimize for the Global South, without compromising experience, so that all users can benefit from the product.
“We’re trying to take our knowledge of technology and UX design, of how to solve problems in general, of how to contextualize things for different kinds of people. We’re trying to build things that improve the lives of people who don’t have access to this technology, where there is a massive scarcity in terms of people working within the machine learning or data science space,” said Nyoni, adding that he hopes his agency’s work is laying the groundwork for more involvement from public health stakeholders like governments. He says the public interest and demand is there, even if the infrastructure isn’t yet.
“It has been said, social entrepreneurship will save Africa.”