Advancements in Surgical Video Analysis

The field of surgical video analysis is witnessing significant advancements with the development of innovative methods for point tracking, disparity estimation, and surgical scene segmentation. Researchers are exploring semi-supervised learning approaches to adapt models to real-world surgical videos, addressing challenges such as domain shift and scarcity of labeled data. Noteworthy papers in this area include SurgTracker, which presents a semi-supervised framework for adapting synthetic-trained point trackers to surgical video, and DGORNet, which refines disparity maps by leveraging monocular depth information. Another notable work is ReSurgSAM2, a two-stage surgical referring segmentation framework that achieves substantial improvements in accuracy and efficiency. MrTrack, an aspiration needle tracker with a mamba-based register mechanism, also demonstrates superior performance in tracking needles with rapid reciprocating motion. These developments highlight the potential for improved surgical quality and patient outcomes through enhanced computer-assisted surgery.

Sources

You Are Your Best Teacher: Semi-Supervised Surgical Point Tracking with Cycle-Consistent Self-Distillation

OPTIKS: Optimized Gradient Properties Through Timing in K-Space

Monocular Depth Guided Occlusion-Aware Disparity Refinement via Semi-supervised Learning in Laparoscopic Images

ReSurgSAM2: Referring Segment Anything in Surgical Video via Credible Long-term Tracking

MrTrack: Register Mamba for Needle Tracking with Rapid Reciprocating Motion during Ultrasound-Guided Aspiration Biopsy

Built with on top of