AI Assistant for Surgical Support & Post-Surgical Insights

The Core Challenge

How do we design AI that supports medical professionals without overstepping boundaries?

Goal

Design an AI assistant to support surgeons during procedures and provide post-surgical insights—without disrupting clinical judgment or workflow.

“The AI should never take over. It should support me, not make decisions for me.”
— Gynecologic Oncology Fellow
 

Designed with AI (and humans)

ChatGPT
Drafted interview questions and generated interaction ideas
Claude
Summarized insights and refined design concepts
Midjourney
Visualized AR overlay concepts for the operating room
Figma
Prototyped interface flows and visual components
Replit
Explored more advanced, interactive prototypes to simulate logic and behavior
UX Pilot
Used to evaluate usability, clarity, and visual hierarchy of early designs
⚕️
Real surgeon feedback
Grounded interaction timing, trust signals, and suggestion tone in clinical expectations
 

Research & Strategic Constraints

3
Research
Read 3 research papers about how AI is used in surgery.
3
Expert Interviews
Talked to 3 surgeons about how they want AI to fit into their workflow
100%
Rules & Safety
Reviewed FDA device guidelines and HIPAA rules.
 

Solution Overview: Two modes of AI assistance

Surgical Vision Modes
Training Mode
Resident Mode Interface
An intelligent guidance system that provides real-time feedback and educational prompts to help surgical residents build confidence and expertise.
  • AI-powered guidance system
  • Real-time feedback and corrections
  • Progress tracking and analytics
Expert Mode
Pro Mode Interface
A streamlined experience that enhances decision-making with subtle prompts and post-op insights tailored for experienced surgeons.
  • Minimal, non-intrusive interface
  • Post-op analytics dashboard
  • Custom alert thresholds
 

Key Design Features

  1. Surgical Procedure Screen

 

2. AI Control Panel

 

3. Post-Op Dashboard & Report

 
 
 

If the project continued, we’d track:

  • Surgeon focus (via eye tracking or NASA-TLX)

  • How often AI prompts were used or skipped

  • Accuracy of key surgical steps

  • Speed of critical decisions

    Let me know if you’d like to see more—I’m happy to share!