Multi-Scale Attention Networks with Feature Refinement for Medical Item Classification in Intelligent Healthcare Systems
- Waqar Riaz 1,2, Asif Ullah 1,2, Jiancheng Charles Ji 1
- Waqar Riaz 1,2, Asif Ullah 1,2, Jiancheng Charles Ji 1
- 1Institute of Intelligent Manufacturing Technology, Shenzhen Polytechnic University, 4089 Shahe West Road, Shenzhen 518055, China.
- 2Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China.
- 0Institute of Intelligent Manufacturing Technology, Shenzhen Polytechnic University, 4089 Shahe West Road, Shenzhen 518055, China.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.This study introduces a hybrid deep learning framework for accurate medical item recognition in healthcare inventory. The AI system achieves high precision in detecting and classifying diverse medical supplies, improving inventory management.
Area Of Science
- Medical Imaging
- Artificial Intelligence
- Computer Vision
Background
- Intelligent healthcare systems require accurate medical imaging for inventory management.
- Efficient recognition and classification of medical items are crucial for inventory integrity and timely access to resources.
Purpose Of The Study
- To present a hybrid deep learning framework, EfficientDet-BiFormer-ResNet, for robust medical item detection and classification in healthcare inventory.
- To enhance the accuracy and efficiency of AI-driven medical inventory solutions.
Main Methods
- Developed a hybrid deep learning framework integrating EfficientDet (BiFPN), BiFormer (attention), and ResNet-18 (triplet loss, OHNM).
- Trained and validated the model on a custom dataset of over 5000 medical inventory images under varied conditions.
- Evaluated performance using mean average precision (mAP) and top-1 classification accuracy.
Main Results
- Achieved a mean average precision (mAP@0.5:0.95) of 83.2% and a top-1 classification accuracy of 94.7%.
- Outperformed conventional models like YOLO, SSD, and Mask R-CNN.
- Demonstrated superior performance in recognizing visually similar, occluded, and small-scale medical items.
Conclusions
- The EfficientDet-BiFormer-ResNet framework provides an effective AI-enabled vision system for real-time medical inventory management.
- This advancement supports clinically relevant and accurate detection of medical items in healthcare settings.
- The proposed system enhances inventory integrity and ensures timely access to critical medical supplies.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.