Related Experiment Video
Updated: May 24, 2025
08:27
A New Hybrid Quantitative Evaluation Model for Axillary Junctional Hemorrhage in Swine
Published on: December 6, 2024
204
Deep Learning-Based Integrated System for Intraoperative Blood Loss Quantification in Surgical Sponges
IEEE Journal of Biomedical and Health Informatics
|March 3, 2025
Summary
MDCare, a deep learning system, precisely quantifies surgical blood loss using sponges. This advanced technology offers real-time data, improving patient safety and surgical outcomes over traditional methods.
Area of Science:
- Medical Technology
- Artificial Intelligence in Surgery
- Surgical Informatics
Background:
- Intraoperative blood loss quantification is vital for patient safety and surgical success.
- Traditional visual estimation methods are subjective and inaccurate.
- There is a need for precise, real-time blood loss monitoring systems.
Purpose of the Study:
- To introduce and validate MDCare, a deep learning-integrated system for accurate blood loss quantification using surgical sponges.
- To demonstrate the superiority of MDCare over traditional estimation techniques.
Main Methods:
- Integration of mass sensor and webcam with deep learning algorithms (ResNet-18, YOLOv4).
- Real-time image processing at 7.4 frames per second.
- Validation using both synthetic and real blood scenarios.
Main Results:
- Achieved up to 96.2% classification accuracy for blood loss.
- Demonstrated over 91% sponge detection accuracy.
- System operates in real-time, suitable for surgical environments.
Conclusions:
- MDCare significantly enhances blood loss estimation accuracy compared to traditional methods.
- The system supports surgeons with real-time data for improved decision-making.
- MDCare represents a significant advancement in surgical care, with potential for widespread adoption.
