Min-Kook Choi
Papers
5
Total Citations
32
H-Index
4
About
Min-Kook Choi is a researcher advancing the field of surgical data science, with a focus on computer vision and multimodal learning for minimally invasive procedures. His work centers on workflow recognition, instrument localization, and intraoperative event detection—critical components for developing context-aware surgical assistance systems. Choi has made notable contributions to the PEg TRAnsfer Workflow Recognition Challenge, where his research demonstrated that integrating multimodal data (e.g., video, kinematic, and audio signals) can improve surgical phase recognition, a finding that has garnered over 12 citations. He also developed the hSDB-instrument database, a specialized resource for instrument localization in laparoscopic and robotic surgeries, cited 7 times for its utility in training and benchmarking detection models. Addressing real-world clinical challenges, Choi proposed a semi-supervised learning approach to handle class imbalance in instrument detection, achieving 5 citations. More recently, he introduced an image segmentation-guided model for intraoperative active bleeding detection, a task complicated by visual similarities between active and non-active bleeding. This work, cited 3 times, holds promise for image-guided surgery and as a prognostic indicator. Through these efforts, Choi is shaping safer, more intelligent surgical environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5