Task 3: MMDental Multimodal Analysis
Task 3 is based on the MMDental dataset described in MMDental: A multimodal dataset of tooth CBCT images with expert medical records (Scientific Data, 2025). It extends STS 2026 from image-only analysis toward multimodal dental AI.
The task uses tooth CBCT images together with expert medical records, including clinical descriptions, initial diagnoses and follow-up documentation. Participants are expected to build methods that connect visual CBCT evidence with structured clinical context for dental diagnosis, reporting and treatment reasoning.
This multimodal task highlights the broader clinical workflow around dental CBCT: not only locating anatomy, but also interpreting patient-level information with expert documentation. Detailed submission requirements and evaluation metrics will be released with the Task 3 protocol.
The reference paper for this task is available here: MMDental dataset paper.
The Task 3 leaderboard and winners will be announced after the final evaluation phase is complete.
The latest public materials for STS 2026 will also be linked from the official GitHub repository.