Video-Based Surgical Tool-Tip and Keypoint Tracking Using Multi-Frame Context-Driven Deep Learning Models

ISBI 2025

Bhargav Ghanekar1, Lianne R. Johnson1, Jacob L. Laughlin1, Marcia K. O'Malley1, Ashok Veeraraghavan1
Rice University, Houston TX USA1
Paper ArXiv Code Dataset
Rice

Principle

We employ multi-frame context by refine keypoint ROI segmentation maps with additional information from previous frames, which helps to improve the tracking performance.

Acknowledgement

This work was supported by Rice-Houston Methodist Seed Grant fund.