文献速递神经MR

认知障碍患者中的运动信息引导三维深度学习重建

Motion-Informed 3D Deep Learning Reconstruction in Patients with Cognitive Impairment.
PubMed 文献 · AJNR Am J Neuroradiol · 2026-02-03

这项前瞻性个体内比较研究纳入健康志愿者控制运动队列和评估记忆丧失的临床患者队列,评估整合回顾性运动校正的深度学习重建在4倍欠采样三维MPRAGE脑MRI中的表现。结果显示,深度学习法在中度和重度运动下显著降低分割误差(分别从12.4%降至3.5%和从44.2%降至12.5%,P < .001),并改善视觉评分(整体图像质量4.26 vs 3.59,P < .001),47%的病例运动伪影严重程度改善。作者认为该技术可能提高诊断效用并降低易运动患者的扫描失败率。

要点

  • 深度学习重建整合运动校正,在中重度运动下分割误差分别从12.4%降至3.5%和从44.2%降至12.5%(P < .001)。
  • 整体图像质量评分深度学习法4.26(SD 0.72)优于标准重建3.59(SD 0.82),P < .001。
  • 47%的病例经深度学习处理后运动伪影严重程度改善,读者间一致性为中等至高度。

核心结论

  • 运动信息引导的深度学习重建改善了三维T1加权脑MRI的形态学准确性和感知图像质量。
  • 该技术可能提高诊断效用并降低认知障碍易运动患者的扫描失败率。

研究设计与样本

  • 前瞻性个体内比较研究,纳入健康志愿者控制运动队列和评估记忆丧失的临床患者队列,各队列在不同成像中心于2022年10月至2023年8月期间分阶段扫描。
  • 共41名参与者(15名女性[37%];平均年龄58岁)和154个图像容积。
  • 所有参与者接受4倍欠采样三维MPRAGE,并整合scout加速运动估计与减少(SAMER)采集。
  • 图像容积分别用标准护理方法和提出的深度学习方法重建;健康志愿者中通过比较指令运动扫描与无运动参考扫描的脑分割结果评估定量形态学准确性。
  • 由2名委员会认证的神经放射学家使用5点Likert量表评定图像质量;统计分析包括Wilcoxon检验和组内相关系数。

诊断标准与关键数值

  • 在健康志愿者中,通过比较指令运动扫描与无运动参考扫描的脑分割结果评估定量形态学准确性。
  • 图像质量由2名委员会认证的神经放射学家使用5点Likert量表评定。
  • 深度学习法在中度运动下分割误差从12.4%降至3.5%,在重度运动下从44.2%降至12.5%(P < .001)。
  • 整体图像质量评分深度学习法4.26(SD 0.72)vs 标准重建3.59(SD 0.82),P < .001。
  • 47%的病例运动伪影严重程度在深度学习处理后改善;读者间一致性为中等至高度。

局限性

  • 原文未报告局限性。

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)。

英文原文(摘要)· 点击展开

BACKGROUND AND PURPOSE: Motion artifacts remain a key limitation in brain MRI, particularly during 3D acquisitions in cognitively impaired patients. Most deep learning (DL) reconstruction techniques improve the SNR but lack explicit mechanisms to correct for motion. This study aims to validate a DL reconstruction method that integrates retrospective motion correction into the reconstruction pipeline for 3D T1-weighted brain MRI.

MATERIALS AND METHODS: This prospective, intraindividual comparison study included a controlled-motion cohort of healthy volunteers and a clinical cohort of patients undergoing evaluation for memory loss. Each cohort was scanned at distinct imaging sites between October 2022 and August 2023 in staggered periods. All participants underwent 4-fold undersampled 3D MPRAGE with an integrated scout accelerated motion estimation and reduction (SAMER) acquisition. Image volumes were reconstructed by using standard of care methods and the proposed DL approach. Quantitative morphometric accuracy was assessed by comparing brain segmentation results of instructed-motion scans with motion-free reference scans in the healthy volunteers. Image quality was rated by 2 board-certified neuroradiologists by using a 5-point Likert scale. Statistical analysis included Wilcoxon tests and intraclass correlation coefficients.

RESULTS: A total of 41 participants (15 women [37%]; mean age, 58 years) and 154 image volumes were evaluated. The DL-based method with integrated motion correction significantly reduced segmentation error under moderate and severe motion (12.4% to 3.5% and 44.2% to 12.5%, respectively; P < .001). Visual ratings showed improved scores across all criteria compared with standard reconstructions (overall image quality, 4.26 [SD, 0.72] versus 3.59 [SD, 0.82]; P < .001). In 47% of cases, motion artifact severity was improved following DL-based processing. Interreader agreement ranged from moderate to substantial.

CONCLUSIONS: Motion-informed DL reconstruction improved both morphometric accuracy and perceived image quality on 3D T1-weighted brain MRI. This technique may enhance diagnostic utility and reduce scan failure rates in motion-prone patients with cognitive impairment.

来源:AJNR Am J Neuroradiol,查看原文。版权归原作者所有。

引用

  1. Fujita S, Polak D, Nickel D, et al. Motion-Informed 3D Deep Learning Reconstruction in Patients with Cognitive Impairment. AJNR Am J Neuroradiol. 2026. PMID: 41062184. doi:10.3174/ajnr.A8977

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。