Multi-Scale Attention Networks with Feature Refinement for Medical Item Classification in Intelligent Healthcare Systems
Sensors · 26.08.2025 · 25 (17) · 5305
DOI: 10.3390/s25175305 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 26.08.2025. Verfasst von Waqar Riaz und 2 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Intelligent Manufacturing Technology, Shenzhen Polytechnic University, 4089 Shahe West Road, Shenzhen 518055, China. Hauptthemen laut MEDLINE: Artificial Intelligence, Delivery of Health Care. Die Arbeit wurde bisher 15-mal zitiert (Crossref) und nennt 52 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: school-level project of SZPU Grants, Post-doctoral Later-stage Foundation Project of Shenzhen Polytechnic University, Shenzhen Polytechnic Scientific Research Start-up Project.
Abstract (Originalsprache)
The increasing adoption of artificial intelligence (AI) in intelligent healthcare systems has elevated the demand for robust medical imaging and vision-based inventory solutions. For an intelligent healthcare inventory system, accurate recognition and classification of medical items, including medicines and emergency supplies, are crucial for ensuring inventory integrity and timely access to life-saving resources. This study presents a hybrid deep learning framework, EfficientDet-BiFormer-ResNet, that integrates three specialized components: EfficientDet’s Bidirectional Feature Pyramid Network (BiFPN) for scalable multi-scale object detection, BiFormer’s bi-level routing attention for context-aware spatial refinement, and ResNet-18 enhanced with triplet loss and Online Hard Negative Mining (OHNM) for fine-grained classification. The model was trained and validated on a custom healthcare inventory dataset comprising over 5000 images collected under diverse lighting, occlusion, and arrangement conditions. Quantitative evaluations demonstrated that the proposed system achieved a mean average precision (mAP@0.5:0.95) of 83.2% and a top-1 classification accuracy of 94.7%, outperforming conventional models such as YOLO, SSD, and Mask R-CNN. The framework excelled in recognizing visually similar, occluded, and small-scale medical items. This work advances real-time medical item detection in healthcare by providing an AI-enabled, clinically relevant vision system for medical inventory management.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute of Intelligent Manufacturing Technology, Shenzhen Polytechnic University, 4089 Shahe West Road, Shenzhen 518055, China.
- Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China.
ORCID-Profile: Waqar Riaz, Asif Ullah, Jiancheng Charles Ji
Förderung
- school-level project of SZPU Grants
- Post-doctoral Later-stage Foundation Project of Shenzhen Polytechnic University
- Shenzhen Polytechnic Scientific Research Start-up Project
- This work is Supported by Post-doctoral Foundation Project of Shenzhen Polytechnic University
- Post-doctoral Foundation Project of Shenzhen Polytechnic University
Zitieren
APA
Riaz, W., Ullah, A., & Ji, J. C. (2025). Multi-Scale Attention Networks with Feature Refinement for Medical Item Classification in Intelligent Healthcare Systems. Sensors, 25(17), 5305. https://doi.org/10.3390/s25175305
Vancouver
Riaz W, Ullah A, Ji JC. Multi-Scale Attention Networks with Feature Refinement for Medical Item Classification in Intelligent Healthcare Systems. Sensors. 2025;25(17):5305. doi:10.3390/s25175305
BibTeX
@article{Riaz2025,
author = {Riaz, Waqar and Ullah, Asif and Ji, Jiancheng Charles},
title = {Multi-Scale Attention Networks with Feature Refinement for Medical Item Classification in Intelligent Healthcare Systems},
journal = {Sensors},
year = {2025},
volume = {25},
number = {17},
pages = {5305},
doi = {10.3390/s25175305}
}
EUVIMED