提出变形运动校正中拟合误差不可靠，需结合其他指标判断。
Is fitting error a reliable metric for assessing deformable motion correction in quantitative MRI?
- 用拟合误差评估心脏弥散张量成像的形变配准效果。
- 拟合误差与负特征值相关，但会受心肌细胞断裂影响而失真。
- 建议结合雅可比行列式等指标综合评估，避免误判。
定量磁共振（qMR）可提供组织物理化学特性的数值表征。为在不同设置下采集一系列帧图像，回顾性运动校正对于对齐对应解剖点或特征至关重要。基于错位会导致对应特征间差异增大的假设，拟合误差常被用作qMR中运动校正的评估指标。本研究评估了在心脏弥散张量成像（cDTI）经形变配准后，拟合误差指标的可靠性。发现尽管拟合误差与负特征值存在相关性，但负雅可比行列式随心肌细胞断裂增加，表现为螺旋角梯度线轮廓异常。由于拟合误差衡量的是移动点与其重渲染对应点之间的距离，其参数本身可能因配准质量差而被调整。因此，拟合误差在形变配准中是必要但非充分的指标，应与其他指标结合使用。
原文摘要 · Abstract (English)
Quantitative MR (qMR) can provide numerical values representing the physical and chemical properties of the tissues. To collect a series of frames under varying settings, retrospective motion correction is essential to align the corresponding anatomical points or features. Under the assumption that the misalignment makes the discrepancy between the corresponding features larger, fitting error is a commonly used evaluation metric for motion correction in qMR. This study evaluates the reliability of the fitting error metric in cardiac diffusion tensor imaging (cDTI) after deformable registration. We found that while fitting error correlates with the negative eigenvalues, the negative Jacobian Determinant increases with broken cardiomyocytes, indicated by helix angle gradient line profiles. Since fitting error measures the distance between moved points and their re-rendered counterparts, the fitting parameter itself may be adjusted due to poor registration. Therefore, fitting error in deformable registration itself is a necessary but not sufficient metric and should be combined with other metrics.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。