Medical Lumbar Spine 3D Axial MRI Dataset for Stenosis Detection, Severity Classification, and Anatomical Segmentation

☐ Dataset Discription: ● The Medical Lumbar Spine 3D Axial MRI Dataset contains 500 anonymized axial lumbar spine MRI examinations from 500 patients acquired at Fırat University Hospital (FUH) under the bilateral protocol of join international research project between Korea and Türkiye. MRI scans were obtained using routine T2-weighted axial sequences on Philips Ingenia (1.5T/3T), GE Signa HDxt (1.5T), and GE Signa Excite (1.5T) scanners. The dataset supports research on lumbar spinal stenosis localization, severity classification, and axial-wise anatomical segmentation. ● The dataset consists of three components: 1. 3D DICOM Volumes: Original 3D axial MRI volumes organized by patient, covering lumbar levels starting L1–L2 to L5–S1 (5 vertebrae). All DICOM files were anonymized to remove patient-identifying information while preserving essential imaging metadata. 2. Localization and Severity Classification: Axial PNG images with Pascal VOC XML annotations. Expert-labeled bounding boxes identify five anatomical regions: Central Canal Stenosis (CCS), Left and Right Lateral Recess Stenosis (LLS and RLS), and Left and Right Foraminal Stenosis (LFS and RFS). Each annotation includes bounding-box coordinates, lumbar level (L1-L2 to L5-S1), AP diameter (mm), anatomical region, and stenosis grade (Normal, Stenosis, or Severe Stenosis). 3. Anatomical Segmentation: Axial MRI slices with pixel-wise masks for Posterior Elements (PE), Central Canal (CC), Area of Anterior–Posterior (AAP), and Intervertebral Disc (IVD). Mask labels are encoded for visualization as: 0 (Background), 50(PE), 100(CC), 150(AAP), and 200(IVD). Masks were generated using AI and refined through manual correction and expert neuroradiologists' validation. ☐ ACKNOWLEDGMENT: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. RS-2023-00256517) and by the TUBITAK (The Scientific and Technological research Council of Turkey) under Grant Number: 123N325. ☐ Please cite these articles when you use our dataset: [1] Abdulmahmod, Osamah F., Mugahed A. Al-antari, Hyunwook Kwon, Afnan Habib, Mukhlis Raza, Metin Kaplan, Bilal Ertuğrul, İsmail Akçin, Ertan Bütün, and Yeong Hyeon Gu. "Medical Spine Sagittal MRI Dataset for Segmentation and Foraminal Stenosis Detection." Scientific Data 13 (2026): Article 809. https://doi.org/10.1038/s41597-026-07138-x [2] Al-Antari, Mugahed A., Saied Salem, Mukhlis Raza, Ahmed S. Elbadawy, Ertan Bütün, Ahmet Arif Aydin, Murat Aydoğan, Bilal Ertuğrul, Muhammed Talo, and Yeong Hyeon Gu. "Evaluating AI-powered predictive solutions for MRI in lumbar spinal stenosis: a systematic review." Artificial Intelligence Review 58, no. 8 (2025): 221.☐ 数据集描述： ● 本医学腰椎3D轴向磁共振成像（Magnetic Resonance Imaging, MRI）数据集包含500例经匿名化处理的腰椎轴向MRI检查数据，来自500名患者，采集自土耳其菲拉特大学医院（Fırat University Hospital, FUH），基于韩土双边联合国际研究项目的协议完成。MRI扫描采用常规T2加权轴向序列，使用Philips Ingenia（1.5T/3T）、GE Signa HDxt（1.5T）及GE Signa Excite（1.5T）扫描仪完成。本数据集可用于腰椎管狭窄症定位、严重程度分级以及轴向解剖分割相关研究。 ● 本数据集包含三个组成部分： 1. 3D医学数字成像与通信（Digital Imaging and Communications in Medicine, DICOM）体积数据：以患者为单位组织的原始3D轴向MRI体积数据，覆盖腰椎节段L1–L2至L5–S1（共5个椎节）。所有DICOM文件均已完成匿名化处理，移除患者识别信息，但保留必要的成像元数据。 2. 定位与严重程度分级数据：带有Pascal VOC格式XML标注的轴向PNG图像。经专家标注的边界框可识别5个解剖区域：中央椎管狭窄（Central Canal Stenosis, CCS）、左侧及右侧侧隐窝狭窄（Left and Right Lateral Recess Stenosis, LLS与RLS）、左侧及右侧椎间孔狭窄（Left and Right Foraminal Stenosis, LFS与RFS）。每份标注包含边界框坐标、腰椎节段（L1-L2至L5-S1）、前后径（mm）、解剖区域以及狭窄分级（正常、狭窄或重度狭窄）。 3. 解剖分割数据：带有像素级掩码的轴向MRI切片，覆盖后部结构（Posterior Elements, PE）、中央椎管（Central Canal, CC）、前后区域（Area of Anterior–Posterior, AAP）以及椎间盘（Intervertebral Disc, IVD）。掩码标签按可视化需求编码为：0（背景）、50（后部结构）、100（中央椎管）、150（前后区域）、200（椎间盘）。掩码通过人工智能（Artificial Intelligence, AI）生成，并经人工校正及神经放射学专家验证优化。 ☐ 致谢：本研究获得韩国国家研究基金会（National Research Foundation of Korea, NRF）资助（由韩国科学和信息通信技术部（Ministry of Science and ICT, MSIT）拨款，编号RS-2023-00256517），以及土耳其科学和技术研究理事会（The Scientific and Technological Research Council of Turkey, TÜBİTAK）资助（编号123N325）。 ☐ 使用本数据集时，请引用以下文献： [1] Abdulmahmod, Osamah F., Mugahed A. Al-antari, Hyunwook Kwon, Afnan Habib, Mukhlis Raza, Metin Kaplan, Bilal Ertuğrul, İsmail Akçin, Ertan Bütün, and Yeong Hyeon Gu. "用于分割及椎间孔狭窄检测的医学脊柱矢状位MRI数据集". 《科学数据（Scientific Data）》13 (2026): 文章809. https://doi.org/10.1038/s41597-026-07138-x [2] Al-Antari, Mugahed A., Saied Salem, Mukhlis Raza, Ahmed S. Elbadawy, Ertan Bütün, Ahmet Arif Aydin, Murat Aydoğan, Bilal Ertuğrul, Muhammed Talo, and Yeong Hyeon Gu. "评估人工智能驱动的腰椎管狭窄症MRI预测方案：系统综述". 《人工智能评论（Artificial Intelligence Review）》58, no. 8 (2025): 221.