DiagnoLens AI

DiagnoLens AI is an AI-powered diagnostic copilot designed to help clinicians detect cognitive bias, uncover missed diagnoses, and recommend critical rule-out tests before it is too late.

Comment

Inspiration

Diagnostic errors are among the leading causes of preventable patient harm. In many cases, clinicians become anchored to the most obvious diagnosis and may overlook less common but life-threatening alternatives.

DiagnoLens AI was created to help physicians look deeper, challenge cognitive bias, and uncover hidden diagnoses before they lead to delayed treatment, unnecessary procedures, or avoidable harm.

What it does

DiagnoLens AI is an AI-powered diagnostic copilot that analyzes patient symptoms, history, and preliminary diagnoses to detect cognitive biases such as anchoring bias and premature closure.

The system expands the differential diagnosis, highlights dangerous alternatives, and recommends critical rule-out tests to ensure clinicians consider what might otherwise be missed.

Key Outputs

  • Anchoring Bias Score
  • Potential Harm Assessment
  • Diagnostic Confidence Score
  • Estimated Time Saved
  • Alternative Diagnoses
  • Missing Rule-Out Tests
  • Physician Summary
  • PDF and JSON Export

How we built it

DiagnoLens AI was built as a full-stack healthcare AI application using:

  • Python
  • FastAPI
  • React.js
  • JavaScript
  • HTML/CSS
  • Prompt Opinion A2A Protocol
  • ReDx AI Model

The backend processes structured clinical case data and performs diagnostic reasoning, while the frontend presents results through a polished dashboard optimized for clinicians.

We deployed the application as a live A2A-enabled BYO Agent on the Prompt Opinion platform and published it to the Prompt Opinion Marketplace.

Challenges we ran into

One of the biggest challenges was creating realistic medical demo cases with meaningful and distinct outputs for each patient scenario.

We also had to:

  • Configure A2A communication and agent skills
  • Publish the agent to the Prompt Opinion Marketplace
  • Design clinically relevant scoring metrics
  • Build an intuitive and professional user interface
  • Ensure the application was fully functional end-to-end within the hackathon timeline

Accomplishments that we're proud of

  • Built a production-quality healthcare AI application
  • Published a working Prompt Opinion Marketplace agent
  • Enabled A2A interoperability with a live agent card
  • Developed realistic diagnostic scenarios
  • Implemented PDF and JSON export functionality
  • Produced a polished demo video and public submission

What we learned

Through this project, we gained valuable experience in:

  • Clinical diagnostic reasoning
  • Cognitive bias detection
  • Healthcare AI product design
  • Full-stack development
  • Prompt Opinion A2A integration
  • Building interoperable agents for real-world healthcare workflows

What's next for DiagnoLens AI

Future improvements include:

  • Real EHR and FHIR integration
  • Explainable reasoning traces
  • Multi-agent specialist collaboration
  • Clinical validation with healthcare professionals
  • Predictive models trained on real-world medical datasets
  • Integration into hospital decision-support systems

Try DiagnoLens AI

Live Agent

https://app.promptopinion.ai/marketplace/agent/019e1c98-41d7-7b33-ae22-7d08aa79778d

A2A Agent Card

https://app.promptopinion.ai/api/workspaces/019e0b84-cb13-7437-b437-58d4f32cd035/ai-agents/019e1241-adaa-7808-b03c-bf7d9765bdd3/.well-known/agent-card.json

Demo Video

https://youtu.be/Vgy1g4_byng

Built With

Share this project:

Updates