Designing Trust: How We Turned Tech into a Doctor’s Ally
In 2019 as the only designer on a team of group of computer scientists and engineers, I found myself at the heart of an innovation hackathon, driven by a mission to tackle a critical societal issue: early diagnosis of skin cancer. Our research revealed a startling fact—doctors were only achieving a 66% accuracy rate in diagnosing skin cancer. This revelation ignited a fire within our team to create a solution that could make a real difference
We decided to harness the power of machine learning, using a dataset of skin samples from real cases and comparing them with images taken on an iPad or tablet. The goal was ambitious, but we were determined. During the prototyping phase, I took on the responsibility of developing a quick prototype and presenting our concept to physicians and healthcare organizations.
However, our initial excitement was met with resistance. The doctors we presented to felt threatened by the technology, fearing it might replace their jobs. This friction was disheartening, but it also made me realize the importance of empathy and understanding in innovation.
Determined to bridge this gap, I went back to the drawing board. I knew we had to present our app as an ally, not a threat. I redesigned the product experience, incorporating an onboarding process for doctors that clearly explained the benefits and limitations of the technology. It was crucial to convey that the model wasn’t 100% accurate and that human expertise was always essential for analysis.
With this new approach, we presented our app once again. This time, the response was different. The doctors saw the potential of the technology as a tool to enhance their diagnostic capabilities, not replace them. Our model, trained on a small dataset, achieved a 66% accuracy rate by analyzing skin samples pixel by pixel and evaluating colors. It allowed doctors to assess the texture and elevation of the area in the sample, providing quick false positive or false negative results. This enabled them to either disregard the probability for further analysis or take it seriously and recommend additional tests to prevent risks.
In the end, our solution placed #2 in the contest. But more importantly, we had created a proof of concept that showed how technology and human expertise could work hand in hand. This journey taught me the power of empathy, collaboration, and perseverance in overcoming challenges and making a meaningful impact.
Comments