PATG: position-aware temporal graph networks for surgical phase recognition on laparoscopic videos.
Abdolrahim KadkhodamohammadiImanol LuengoDanail StoyanovPublished in: Int. J. Comput. Assist. Radiol. Surg. (2022)
Keyphrases
- robot assisted
- human activities
- recognition rate
- laparoscopic surgery
- motion history images
- temporal segmentation
- spatio temporal
- minimally invasive surgery
- operating room
- temporal information
- recognition algorithm
- object recognition
- small world
- social networks
- temporal relationships
- average degree
- fully connected
- temporal domain
- surgical procedures
- temporal coherence
- feature extraction
- temporal constraints
- intraoperative
- video sequences
- random walk
- graph structure
- space time
- community discovery
- edge weights
- spatial and temporal
- directed graph
- graph structures
- graph theory
- graph layout
- network structure
- complex networks
- dynamic textures
- graph mining
- temporal data
- bipartite graph
- degrees of freedom
- minimally invasive
- human actions
- master slave
- video frames
- activity recognition
- video data
- medical images
- weighted graph
- surgical robot