We are thrilled to share that our international research collaboration between China, Iran, the USA, and Italy has led to an outstanding achievement!
Our paper, titled:
“Unfiltered Sonographic Images for Kidney Study: Evaluating YOLO Variants with Attention Mechanisms for Anatomical Segmentation”
has been selected as the Best Paper Award Winner at the 🏆 IEEE 3rd International Conference on Artificial Intelligence and Automation Control (AIAC 2025),held October 15–17, 2025, in Paris, France.
🧠 Research Overview
Ultrasound imaging is a key tool in kidney studies because of its affordability, accessibility, and non-invasive nature. However, challenges such as noise, low contrast, and imaging artifacts in raw sonographic data limit its usefulness for AI-based anatomical analysis.This study investigates the performance of three YOLO model variants—each integrated with advanced attention mechanisms—using a real-world, unfiltered kidney ultrasound dataset with expert-verified annotations from both male and female patients.Performance was evaluated using inference time and segmentation accuracy across 11 anatomical structures. The models showed strong detection in clearly defined regions such as the renal cortex and sinus, while more ambiguous areas like the medulla and Column of Bertin remained challenging due to low contrast.The research establishes an important benchmark for analyzing noisy ultrasound data and highlights the need for customized preprocessing filters to enhance visibility and segmentation in AI-assisted kidney diagnostics.
👥 Authors
Ata Jahangir Moshayedi, School of Information Engineering, Jiangxi University of Science and Technology, Ganzhou, Jiangxi, China
Mohammad Jalil Jawadi, School of Information Engineering, Jiangxi University of Science and Technology, Ganzhou, Jiangxi, China
Samira Hemmati, Radiology Department, Kosar Hospital, Semnan University of Medical Sciences, Semnan, Iran
David Bassir, Smart Structural Health Monitoring and Control Laboratory, DGUT-CNAM, Dongguan University of Technology, Dongguan, China
Tahmineh Mokhtari, Department of Entomology and Nematology and UCD Comprehensive Cancer Center, University of California, Davis, USA
Davood Arab, Clinical Research Development, Kosar Hospital, Semnan University of Medical Sciences, Semnan, Iran
Jianbing Yi, School of Information Engineering, Jiangxi University of Science and Technology, Ganzhou, Jiangxi, China
Mehran Emadi Andani, Department of Neuroscience, Biomedicine and Movement Sciences, University of Verona, Verona, Italy
This achievement marks another milestone in AI-driven medical imaging research, showcasing the potential of deep learning and attention mechanisms to deliver accurate, efficient, and interpretable kidney ultrasound diagnostics.
