DermDetect: Advancing Dermatological Accessibility
Role: UX Researcher & Designer
Teammate: K. Castillo
THE CHALLENGE
Imagine having a painful skin condition but no way to see a doctor or afford treatment. That’s the reality for a lot of people in underserved regions. Without the right equipment or local medical pros, simple skin issues can turn into permanent, life-altering conditions. We needed a way to bring the clinic directly to the patient.
THE SOLUTION
We built a prototype app called DermDetect. It’s simply a pocket medical database powered by AI. Users can scan a skin concern, get a potential diagnosis and treatment plan, and even connect with local doctors or community members for support.
HOW WE RESEARCHED:
We kicked off the process by reading existing research and writing up a literature review. We put together a quick synopsis for every article we collected to really understand the current landscape of dermatological apps. When it came to early designs, we wanted the focus to stay completely on the layout and user flow. So, all the initial wireframes were hand-drawn and then created in Figma. We then built the app in Adalo.
We designed DermDetect with our specific users in mind, meaning we had to think about hurdles like spotty internet connections. We wanted a "snap & scan" capability where users take a picture of their skin issue, and the AI steps in to suggest what it might be and how to treat it. It would even save to a "My Skin History" log so they can look back at past diagnoses. Bad Wi-Fi? No problem. The app would work offline and translate into different languages to fit the user's background. We also wanted a simple scheduling interface to book virtual or in-person visits with nearby pros. A built-in forum would let users ask questions and share experiences with folks in their area. Users can tweak everything from captions to notification reminders and privacy settings for sharing data with doctors.
We reached out to about 100 people using systematic sampling, getting feedback from locals in Hudson County and the general public.
Participants interacted with the Adalo prototype by the tasks we gave them. We then asked them to take a survey to evaluate the app's functionality, engagement, information quality, and aesthetics on a scale of 1 to 5.
PERSONA & PROBLEM STATEMENTS:
BUILDING THE DERMDETECT APP
First, we drew our ideas out on paper.
Then, we used Figma to design the interface and map out the user flow. We then built the DermDetect app in Adalo.
USER FLOW
HOW THE DERMDETECT APP WORKS
DermDetect turns your smartphone into a pocket-sized skin specialist.
You just snap a quick photo of whatever skin issue is bothering you, and the app uses AI to analyze the image and figure out what might be going on.
It instantly gives you potential diagnoses alongside treatment ideas.
It can even connect you with nearby doctors for a quick virtual chat if you need a professional opinion
LIVE PROTOTYPE CREATED ON ADALO
USABILITY TESTING
Think-Aloud Cognitive Walkthrough
Task 1. You have a new spot on your arm that you want to check. Talk me through how you would navigate from the home screen to open the scanner and capture a photo.
Task 2. You just took a picture of your skin, but the lighting is bad and it looks blurry. Walk me through how you would discard that image, retake it, and then confirm the final version for analysis.
Task 3. You want to review a diagnosis you received last month. Show me how you would find your past scans and open the specific details for 'Rosacea'.
Task 4. You heard about a condition called "Scabies" and want to know what it looks like. Show me how you would search the app's database to find images and information about it.
Task 5. You are currently looking at a specific skin condition in the search database, but you want to go back to the main starting screen. Talk me through what you would click to get there.
INSIGHTS FROM USER TESTING
-
About 45% of our testers found the app entertaining and effective for short bursts of time.
-
Around 40% of users felt the medical content and AI insights were highly relevant to their needs.
-
A combined majority of our testers said they would recommend the app to several or many people who could benefit from it.
-
Over half the users thought the app was perfectly or well-targeted for people lacking healthcare access.
-
A small group of about 7.5% found the app confusing to use, which tells us we need to clean up the user flow and onboarding process.
-
To truly work in resource-poor areas, users absolutely rely on the offline mode and translation tools to overcome spotty internet and language barriers.
-
Testing showed we need to add a GPS feature to help people find local pharmacies for their treatments.
-
We realized the app needs a solid backend so physicians can securely save patient info and manage their appointments.
VALUE FLOW MAP
