Hierarchical Curriculum Learning for Surgical Video Generation: Focusing on Bilateral Axillo-Breast Approach Robotic Thyroidectomy Datasets

Key points

• High-fidelity surgical video generation is increasingly important for surgical education, simulation, and downstream AI training, as conventional simulation pipelines are limited by cost, realism, and scenario diversity.1
• Adapting generic text-to-video diffusion models to surgical procedures remains difficult, however, due to complex anatomical structures, intricate tool-tissue interactions, the inherently hierarchical nature of surgical workflows, and the scarcity of curated training data outside a few well-studied procedures.1
• Surgical procedures are decomposed into three levels (phase, task, and action) and a pretrained text-to-video diffusion model is progressively adapted from coarse to fine-grained surgical context, supported by a dynamic sampling strategy that handles variable-resolution and variable-length surgical clips.1
• We instantiate the framework on Bilateral Axillo-Breast Approach robotic thyroidectomy, a procedure for which no curated benchmark currently exists.1

Sources (1)

• [1]Hierarchical Curriculum Learning for Surgical Video Generation: Focusing on Bilateral Axillo-Breast Approach Robotic Thyroidectomy Datasets
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 13, 04:07 AM

  We propose a surgical video generation framework that combines hierarchical procedural annotation with curriculum transfer learning.

  High-fidelity surgical video generation is increasingly important for surgical education, simulation, and downstream AI training, as conventional simulation pipelines are limited by cost, realism, and scenario diversity.

Extractive summary: sentences quoted from the sources.