Related Experiment Video
Updated: Jul 9, 2025
07:12
Real-Time Monitoring of Neurocritical Patients with Diffuse Optical Spectroscopies
Published on: November 19, 2020
2.2K
CLAD-Net: cross-layer aggregation attention network for real-time endoscopic instrument detection
Xiushun Zhao1, Jing Guo1, Zhaoshui He1
1School of Automation, Guangdong University of Technology, Guangzhou, 510006 China.
Health Information Science and Systems
|November 29, 2023
Summary
This study introduces CLAD-Net, a novel deep learning model for precise endoscopic instrument detection in minimally invasive surgery (MIS). CLAD-Net significantly improves surgical safety and efficiency by accurately identifying instruments in challenging visual conditions.
Area of Science:
- Medical Imaging
- Computer Vision
- Surgical Technology
Background:
- Minimally invasive surgery (MIS) relies on accurate identification of surgical instruments for effective guidance.
- Endoscopic instrument detection is challenging due to confined spaces, occlusions, and variable lighting.
Purpose of the Study:
- To develop an accurate and real-time detection network for endoscopic instruments in complex MIS scenarios.
- To enhance surgical efficiency and patient safety through improved instrument recognition.
Main Methods:
- Proposed a cross-layer aggregated attention detection network (CLAD-Net).
- Introduced a cross-layer aggregation attention module for enhanced feature fusion and propagation.
- Developed a composite attention mechanism (CAM) for multi-scale contextual information extraction and feature refinement.
- Implemented a feature refinement module (RM) to improve edge and detail extraction.
Main Results:
- CLAD-Net achieved high detection accuracy, with 98.9% on the Cholec80 dataset and 98.6% on a neuroendoscopic dataset.
- The proposed network outperformed existing advanced detection networks in complex surgical environments.
- Demonstrated real-time detection capabilities.
Conclusions:
- CLAD-Net offers a robust solution for accurate endoscopic instrument detection in MIS.
- The proposed attention mechanisms effectively address challenges like inconsistent target size and low contrast.
- This technology has the potential to significantly advance image-guided surgery and improve patient outcomes.
