深度学习自动检测中轴型脊柱关节炎骶髂关节MRI活动性与结构性病变
本研究开发并验证一套全自动深度学习流程,用于在轴型脊柱关节炎(axSpA)患者的骶髂关节(SIJ)MRI中检测1类活动性病变(骨髓水肿)和4类结构性病变(侵蚀、脂肪病变、硬化、强直)。模型在MEASURE 1试验(132例)训练,并在PREVENT(555例)与SURPASS(414例)两个独立数据集外部验证,结果显示自动化检测与专家评估的一致性可与专家间一致性相比,结构性病变表现最佳,且无需额外训练即可跨数据集泛化。
要点
- 全自动深度学习流程可检测骶髂关节MRI的骨髓水肿及侵蚀、脂肪病变、硬化、强直5类病变。
- 强直检测表现最佳(MEASURE 1 AUC 0.97,SURPASS AUC 0.99),骨髓水肿AUC 0.85–0.93但平衡准确度较低(0.72–0.74)。
- 模型在独立试验数据集中外部验证表现稳健,支持用于提升临床试验与观察性研究中MRI评估的一致性与可扩展性。
核心结论
- 全自动深度学习方法可可靠检测axSpA骶髂关节MRI的活动性与结构性病变,并具稳健的外部验证表现。
研究设计与样本
- 开发两阶段自动化流程,分别勾画左右骶髂关节并检测5类MRI病变,按象限或关节水平结合T1加权与短Tau反转恢复序列评估。
- 模型在MEASURE 1试验(132例)以共识标签训练,并在PREVENT(555例)和SURPASS(414例)两个独立数据集验证。
- 性能以AUC、平衡准确度、敏感度、特异度与kappa评估,并对照按柏林骶髂关节评分法的专家评估解读。
主要发现
- 结构性病变表现最强,强直在MEASURE 1的AUC为0.97、SURPASS为0.99,平衡准确度分别为0.95与0.97。
- 侵蚀与脂肪病变亦获稳健结果,骨髓水肿检测AUC持续较高(0.85–0.93),但平衡准确度较低(0.72–0.74)。
- 模型性能跨数据集泛化,无需额外训练。
关键数值与推荐
- 强直:MEASURE 1 AUC 0.97、SURPASS AUC 0.99;平衡准确度0.95与0.97。
- 骨髓水肿:AUC 0.85–0.93,平衡准确度0.72–0.74。
- 检测病变共5类:1类活动性病变为骨髓水肿,4类结构性病变为侵蚀、脂肪病变、硬化、强直。
临床怎么用
- 该自动化流程有望在临床试验与观察性研究中提升骶髂关节MRI评估的一致性与可扩展性,辅助标准化判读。
- 结构性病变(尤其强直)自动检测可靠性较高,骨髓水肿判读平衡准确度相对偏低,临床解读需结合专家评估。
局限性
- 研究基于3项III期临床试验数据集,外部验证限于临床试验人群,未涉及真实世界或其他影像协议。
- 骨髓水肿检测的平衡准确度相对较低(0.72–0.74),提示活动性病变自动判读仍有提升空间。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVES: This study aims to assess the performance of a fully automated deep learning (DL) system for detecting active and structural magnetic resonance imaging (MRI) lesions of the sacroiliac joints (SIJs) in axial spondyloarthritis (axSpA), and validate its generalisability across independent clinical trial datasets. METHODS: A 2-stage automated pipeline was developed to delineate left and right SIJs and detect 5 MRI-defined lesion types: 1 active lesion: bone marrow oedema (BMO), 4 structural lesions: erosions, fat lesions, sclerosis, and ankylosis. Lesions were assessed at the quadrant or joint level using paired T1-weighted and Short Tau Inversion Recovery sequences. Models were trained on the MEASURE 1 trial (132 patients) using consensus-based labels to address multireader variability and evaluated on 2 independent datasets (PREVENT-555 patients; SURPASS-414 patients). Performance was assessed using area under the curve (AUC), balanced accuracy, sensitivity, specificity, and kappa, and interpreted against expert reader evaluations, done according to the Berlin SIJ scoring method. RESULTS: Across all datasets, automated SIJ lesion detection against expert evaluations achieved performance comparable with expert interreader agreement. Structural lesions showed the strongest performance, particularly ankylosis (MEASURE 1 AUC: 0.97, SURPASS AUC: 0.99; balanced accuracy: 0.95 and 0.97). Robust results were also observed for erosions and fat lesions. BMO detection showed consistently high AUCs (0.85-0.93) with lower balanced accuracy (0.72-0.74). Model performance generalised across datasets without additional training. CONCLUSIONS: A fully automated DL-based approach can reliably detect active and structural SIJ MRI lesions in axSpA with robust external validation, supporting its potential use to enhance the consistency and scalability of MRI assessment in clinical trials and observational studies.
引用
- Jamaludin A, Windsor R, Ather S, et al. Automated detection of active and structural MRI lesions in the sacroiliac joints in axial spondyloarthritis: training and validation across 3 phase III clinical trial datasets. Ann Rheum Dis. 2026. PMID: 42701084. doi:10.1016/j.ard.2026.08.003
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。