How It Works
A step-by-step guide to training with LAByrinth and HNLE
Getting Started
Select a case narrative from the home page and follow the guided surgical progression. Each case represents a patient with a clinical story — you'll drill through escalating decision points, making the same decisions you would face in the operating room.
LAByrinth works with both cadaveric temporal bones and 3D printed models, in any lab setting. If you're interested in 3D printed bones, see the Resources page for vendors.
AI-Guided Assessment
At each surgical checkpoint, you can upload a still photo of your dissection — taken with a cell phone camera, microscope camera, USB camera, or other imaging modality. HNLE — the Heuristic Neurotologic Learning Environment — analyzes your image by comparing it against expert reference standards for that checkpoint.
You'll receive structured feedback identifying what was done well and what needs improvement. HNLE acts as a stage gate: you must demonstrate proficiency at each checkpoint before advancing to the next phase of the case.
How HNLE Is Trained
HNLE learns from two complementary sources. First, expert surgeons perform dissections and provide narrative input on landmarks, decision points, and technique. These are stored as gold-standard references. Second, annotated learner mistakes (violated structures, thermal injury, inadequate exposure) teach HNLE what “wrong” looks like.
Instructor annotations go through multi-reviewer feedback and correction before being incorporated. This means the AI continuously improves as more training data is collected and validated.
Disputing Feedback
If you believe HNLE's assessment was inaccurate, you can and should dispute the feedback directly from the results page. Provide a brief explanation of your disagreement, and your dispute will be reviewed by an instructor.
Instructor review may result in the assessment being upheld or corrected. Every dispute is a learning opportunity — both for you and for the AI. You can track your disputes and their outcomes from the “My Feedback” section in the navigation.
Training Analytics
The “My Training” dashboard gives trainees a comprehensive view of their progress: pass rates by difficulty, error type breakdowns, checkpoint coverage, and the ability to revisit and reflect on every prior dissection.
The Antrum provides additional administrative tools for reviewing trainee performance across the program. The administrative team exercises ongoing review and oversight of all analytics, assessment quality, and AI behavior to ensure the system remains accurate and fair.