Redesigning K Health's pediatric Body Picker through usability research, iterative testing and close collaboration with clinical and technical teams.
A parent reports that their child has a rash. The next question sounds simple: Where is it?
In telemedicine, a Body Picker turns what a parent sees into structured medical information. The location may directly affect the clinical assessment.
Specific body parts required an extra zoom step, adding friction to a simple task.
A selection such as “hand” could return as “arm”, leaving users unsure what the system had recorded.
Editing was tedious, the visuals felt overly clinical, and the interaction did not meet the intended usability and accessibility standards.
Help parents confidently describe where a symptom appears, while still collecting clinically meaningful and structured data.
I began with comprehensive competitor research to understand how other medical symptom-based apps approach the diagnostic variable of ‘location’.
Having analyzed solutions from leading apps focusing on usability, accessibility and visual design, I identified industry standards, best practices and common pitfalls, and gathered insights for a better experience.
I then mapped out user pain points through direct feedback and usability studies, and analyzed UX patterns.
The central part of the case was not the first redesign. It was what happened when the prototype met real behavior.
I ran remote sessions in Lookback and observed how parents interpreted the new pattern, switched between body areas and recovered when a part was difficult to find.
The direction worked, while testing exposed a new tension: visual simplicity could make clinically important details harder to find.
Free-form selection opened an important question with developers and data scientists: how much freedom could the interface offer without weakening the clinical ontology and data structure?
Use fewer, broader hit areas to keep the first view calm.
Separate clinically meaningful parts when precision matters.
Provide specific alternatives for details that are difficult to show directly.
I refined the hit areas and visual hierarchy, then ran a second testing round to check whether the changes resolved the issues from Round 1.
For handoff, I defined each illustrated layer as a touchpoint or zoom trigger and mapped it to the correct clinical ontology. This created a precise bridge between the visual interaction and the underlying data.
The second testing round confirmed the refined hierarchy, easier selection of specific body parts and fewer points of hesitation. The final direction combined direct interaction, precise tags and a clinical structure that could be prepared for implementation at scale.
Improving an existing product is not about replacing everything. It is about finding where people lose confidence, testing a clearer direction and staying open when the evidence challenges the first solution. The strongest result came from balancing empathy for the user with the clinical and technical structure the product needed.