: The researchers created a specialized dataset featuring 4,003 hand-labeled frames from laparoscopic videos, including the "VID-0011-1" sequence, to train and test their models. Model Performance :
: Identified as the best compromise, achieving an Intersection over Union (IoU) of 0.85 while running at over 30 frames per second (FPS), making it suitable for live surgical use. VID - 0011-1.mp4
Gauze Detection and Segmentation in Minimally Invasive Surgery Video Using Convolutional Neural Networks : The researchers created a specialized dataset featuring
The video file is a specific sample from the Gauze Detection and Segmentation dataset used in surgical computer vision research. The primary academic paper associated with this video is: The primary academic paper associated with this video
: Miscounting gauze is a common human error in surgery; this paper proposes an automated AI system to track gauze in real-time using laparoscopic camera feeds.
: Capable of real-time detection but had lower recall (missed some gauze).
Published in July 2022, this study addresses the critical medical challenge of —specifically surgical gauze left inside patients after laparoscopic procedures. Key Findings of the Paper: