Surgical Tool Detection
Surgical tool detection in videos aims to automatically identify and locate surgical instruments during procedures, improving analysis and potentially automating aspects of surgery. Current research focuses on overcoming challenges like imbalanced datasets, similar tool appearances, and occlusions, employing techniques such as semi-supervised learning, domain adaptation, and novel loss functions within deep learning models (e.g., transformers). This work is significant for advancing computer-assisted surgery, enabling improved surgical training, objective performance assessment, and ultimately, safer and more efficient procedures.
Papers
October 9, 2024
June 5, 2024
January 5, 2024
September 18, 2023
July 31, 2023
July 3, 2023
May 19, 2023
August 21, 2022