Deep Homography Estimation in Dynamic Surgical Scenes for Laparoscopic Camera Motion Extraction
This addresses the challenge of camera motion automation in laparoscopic surgery, offering a more accurate and efficient alternative to rule-based methods, though it is incremental by building on existing deep learning techniques.
The paper tackles the problem of automating laparoscopic camera motion by introducing a method to extract a laparoscope holder's actions from surgical videos, using synthetic camera motion as a supervisory signal. It outperforms classical homography estimation approaches by 41% in precision and 43% in runtime on a CPU.
Current laparoscopic camera motion automation relies on rule-based approaches or only focuses on surgical tools. Imitation Learning (IL) methods could alleviate these shortcomings, but have so far been applied to oversimplified setups. Instead of extracting actions from oversimplified setups, in this work we introduce a method that allows to extract a laparoscope holder's actions from videos of laparoscopic interventions. We synthetically add camera motion to a newly acquired dataset of camera motion free da Vinci surgery image sequences through a novel homography generation algorithm. The synthetic camera motion serves as a supervisory signal for camera motion estimation that is invariant to object and tool motion. We perform an extensive evaluation of state-of-the-art (SOTA) Deep Neural Networks (DNNs) across multiple compute regimes, finding our method transfers from our camera motion free da Vinci surgery dataset to videos of laparoscopic interventions, outperforming classical homography estimation approaches in both, precision by 41%, and runtime on a CPU by 43%.