发表机构
Business AI Lab, College of Technology, National Economics University; A2I Lab, Phenikaa School of Computing, Phenikaa University; Radiology and Functional Exploration Center, Phenikaa University Hospital（商业人工智能实验室，技术学院，越南国民经济大学; A2I实验室，Ph Phenikaa大学计算学院; 放射学与功能探索中心，Phenikaa大学医院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对脊柱病理学诊断自动化系统缺乏高质量基准的问题，提出PhenSPINE数据集及融合位置编码机制的诊断基准，通过多序列评估发现矢状面T2加权序列诊断价值最佳，建立了基线并为序列选择提供见解。
AI 中文摘要
脊柱疾病的准确诊断很大程度上依赖于放射学解释，但自动化系统因缺乏多样、高质量的基准而受阻。本研究提出PhenSPINE，这是一个包含250名患者的16813张图像的磁共振成像数据集，旨在促进先进的深度学习研究。我们提出了一个强大的诊断基准，将先进的卷积主干与位置编码机制相结合，以明确模拟椎间盘的解剖背景。通过对四个标准MRI序列进行评估，实验表明矢状面T2加权序列具有最强的诊断价值，宏观F1分数达到50.31%。我们发现多序列融合策略的性能不如单序列基线，因为数据集中不同序列的图像受到周围解剖区域噪声干扰的严重影响。这项工作建立了一个强大的基线，并为脊柱分析的序列选择提供了关键见解。
英文摘要
The accurate diagnosis of spinal pathologies depends heavily on radiological interpretation, yet automated systems are hindered by the lack of diverse, high-quality benchmarks. In this study, we present PhenSPINE, a Magnetic Resonance Imaging dataset comprising 16,813 images from 250 patients, curated to facilitate advanced deep learning research. We propose a robust diagnostic benchmark that integrates state-of-theart convolutional backbones with a Positional Encoding mechanism to explicitly model the anatomical context of intervertebral discs. Evaluating across four standard MRI sequences, our experiments demonstrate that the Sagittal T2-weighted sequence offers the most robust diagnostic value, achieving a superior Macro F1-score of 50.31%. We find that multisequence fusion strategies yield inferior performance compared to this single-sequence baseline, as the images across sequences in our dataset are significantly compromised by noise interference from surrounding anatomical regions. This work establishes a robust baseline and offers critical insights into sequence selection for spine analysis.
Comments12 pages, figures, Accepted at CITA 2026 (The 15th Conference on Information Technology and its Applications, Scopus)
