引入级联分割流水线进行心血管磁共振参数 T1 映射，以提高分割性能。

Introduction of a cascaded segmentation pipeline for parametric T1 mapping in cardiovascular magnetic resonance to improve segmentation performance.

机构信息

ECRC Experimental and Clinical Research Center, Charité - Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Lindenberger Weg 80, 13125, Berlin, Germany.

Working Group on Cardiovascular Magnetic Resonance, Experimental and Clinical Research Center, a joint cooperation between the Charité - Universitätsmedizin Berlin and the Max-Delbrück-Center for Molecular Medicine, Berlin, Germany.

出版信息

Sci Rep. 2023 Feb 6;13(1):2103. doi: 10.1038/s41598-023-28975-5.

DOI:10.1038/s41598-023-28975-5

PMID:36746989

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC9902617/

Abstract

The manual and often time-consuming segmentation of the myocardium in cardiovascular magnetic resonance is increasingly automated using convolutional neural networks (CNNs). This study proposes a cascaded segmentation (CASEG) approach to improve automatic image segmentation quality. First, an object detection algorithm predicts a bounding box (BB) for the left ventricular myocardium whose 1.5 times enlargement defines the region of interest (ROI). Then, the ROI image section is fed into a U-Net based segmentation. Two CASEG variants were evaluated: one using the ROI cropped image solely (cropU) and the other using a 2-channel-image additionally containing the original BB image section (crinU). Both were compared to a classical U-Net segmentation (refU). All networks share the same hyperparameters and were tested on basal and midventricular slices of native and contrast enhanced (CE) MOLLI T1 maps. Dice Similarity Coefficient improved significantly (p < 0.05) in cropU and crinU compared to refU (81.06%, 81.22%, 72.79% for native and 80.70%, 79.18%, 71.41% for CE data), while no significant improvement (p < 0.05) was achieved in the mean absolute error of the T1 time (11.94 ms, 12.45 ms, 14.22 ms for native and 5.32 ms, 6.07 ms, 5.89 ms for CE data). In conclusion, CASEG provides an improved geometric concordance but needs further improvement in the quantitative outcome.

摘要

心血管磁共振中手动且耗时的心肌分段正越来越多地使用卷积神经网络（CNN）实现自动化。本研究提出了一种级联分割（CASEG）方法，以提高自动图像分割的质量。首先，目标检测算法预测左心室心肌的边界框（BB），其 1.5 倍放大定义了感兴趣区域（ROI）。然后，将 ROI 图像部分馈送到基于 U-Net 的分割中。评估了两种 CASEG 变体：一种仅使用 ROI 裁剪图像（cropU），另一种使用另外包含原始 BB 图像部分的 2 通道图像（crinU）。将这两种方法与经典的 U-Net 分割（refU）进行了比较。所有网络都共享相同的超参数，并在基底部和中部心室的天然和对比增强（CE）MOLLI T1 图上进行了测试。与 refU 相比，cropU 和 crinU 的 Dice 相似系数显著提高（p < 0.05）（天然数据为 81.06%、81.22%、72.79%，CE 数据为 80.70%、79.18%、71.41%），而 T1 时间的平均绝对误差（11.94 ms、12.45 ms、14.22 ms 用于天然数据和 5.32 ms、6.07 ms、5.89 ms 用于 CE 数据）没有显著提高（p < 0.05）。总之，CASEG 提供了改进的几何一致性，但在定量结果方面需要进一步改进。