MultiBoneX.zip

The <b>MultiBoneX ver1 dataset</b> is a curated collection of 22<b>,270 high-quality X-ray images</b> covering <b>11 anatomical classes</b> of human upper and lower limbs for bone fracture detection. The dataset was constructed by integrating multiple public sources (MURA, FrcAtlas, LERA, HBFMID) and processed through a multi-stage pipeline including data cleaning, annotation verification by expert radiologists, preprocessing, enhancement, augmentation, and class balancing. Each image is labeled hierarchically by both <b>skeletal region</b> and <b>fracture status (fractured/normal)</b>, enabling use in both <b>binary</b> and <b>multi-class classification tasks</b>.The dataset reflects <b>real-world radiology variability</b>, with images sourced from different X-ray machines and conditions to ensure robustness across heterogeneous clinical settings. It provides comprehensive coverage of bone structures such as elbow, forearm, wrist, hand, humerus, hip, knee, ankle, and foot.<b>MultiBoneX ver1</b> is intended to support research in <b>deep learning, medical imaging, and computer-aided diagnosis</b>, and is suitable for developing clinically relevant, generalizable fracture detection systems for both educational and diagnostic applications.MultiBoneX ver1数据集（MultiBoneX ver1 dataset）是一套精心整理的数据集，包含22270张高质量X射线图像，覆盖用于骨折检测的人类上下肢11个解剖类别。该数据集通过整合MURA、FrcAtlas、LERA、HBFMID等多个公开数据源构建而成，并经过多阶段处理流程，涵盖数据清洗、资深放射科医师标注核验、预处理、图像增强、数据扩增以及类别平衡。每张图像均按照骨骼区域与骨折状态（骨折/正常）进行分层标注，可支持二分类及多类别分类任务的开展。该数据集反映了真实临床放射场景中的成像变异性，图像采集自不同型号的X射线设备与成像环境，以确保模型在异质临床场景下的鲁棒性。该数据集全面覆盖了肘关节、前臂、腕关节、手部、肱骨、髋关节、膝关节、踝关节以及足部等骨骼结构。MultiBoneX ver1数据集旨在支持深度学习、医学影像与计算机辅助诊断领域的研究，适用于开发兼具临床实用性与泛化能力的骨折检测系统，可应用于教学与辅助诊断相关场景。