tsta

فارسی

Home / سی و دومین کنفرانس ملی و دهمین کنفرانس بین المللی مهندسی زیست پزشکی ایران

Automated Tibial Bone Segmentation using 2D Swin-Unet on Knee X-ray Images

Authors :

Ali Kazemi¹ Abolfazl Zamanirad² Soodabeh Esfandiary³ Ebrahim Najafzadeh⁴ Mohammad Hossein Nabian⁵ Parastoo Farnia⁶ Alireza Ahmadian⁷

1- دانشگاه علوم پزشکی تهران 2- دانشگاه علوم پزشکی تهران 3- دانشگاه علوم پزشکی تهران 4- دانشگاه علوم پزشکی ایران 5- دانشگاه علوم پزشکی تهران 6- دانشگاه علوم پزشکی تهران 7- دانشگاه علوم پزشکی تهران

Keywords :

Tibial Plateau Fracture،Medical Image Segmentation،Swin-Unet،X-ray Imaging،Deep Learning

Abstract :

Tibial plateau fractures (TPFs), which account for approximately 1% of all bone fractures, represent a complex subset of knee injuries with significant clinical implications if not accurately diagnosed and managed. The accurate diagnosis of TPFs from radiographs is challenged by subtle fracture lines and significant inter-observer variability in manual segmentation. To address these limitations, this study evaluates the performance of a Transformer-based deep learning model, Swin-Unet, for automated and precise tibial segmentation. A retrospective dataset comprising 958 anterior-posterior and lateral radiographs from 220 patients with TPFs was curated. Ground truth masks of the tibia bone were manually annotated and validated through a multi-stage review by orthopedic surgeons. Following preprocessing steps, including contrast enhancement with Contrast Limited Adaptive Histogram Equalization (CLAHE), a 2D Swin-Unet architecture featuring patch-based self-attention mechanisms was trained. The optimized Swin-Unet model demonstrated high fidelity, achieving a mean Dice Similarity Coefficient (DSC) of 0.8314, a mean Intersection over Union (IoU) of 0.7374, and an overall accuracy of 0.9735 on the validation set. Qualitative analysis confirmed the model's ability to accurately delineate tibial boundaries. In conclusion, this study validates the Swin-Unet model as a robust and efficient framework for automated tibial segmentation. By mitigating the challenges of manual delineation, this approach holds significant promise for improving the consistency of orthopedic diagnostic workflows. It serves as a foundation for AI-driven clinical decision support in musculoskeletal imaging.