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

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.

Post-hoc standardisation of parametric T1 maps in cardiovascular magnetic resonance imaging: a proof-of-concept

“… 33 While reconstruction algorithms are not available from DICOM tags, the segmentation method of the myocardium is still based on expert agreement or in accordance to guidelines. 34 Although fully automated segmentation procedures exist, 16 manual adaptions are still necessary. Consequently, the segmentation method is currently a local specific, hard to definable CP.…”

“…However, in future perspective, when including more training data, this must be considered, especially since artificial intelligence assisted segmentation procedures become ever more popular. 16 , 35…”

“…The segmentation of all parametric T1 maps was performed automatically with a subsequent visual inspection by two experts (JG and EB). First, all T1 maps were segmented using the cropU cascaded model 16 and a research deep learning segmentation model provided by Siemens Healthcare (version 21 hotfix, Siemens Healthcare GmbH, Erlangen, Germany). The expert chose the best segmentation out of either model or the intersection of both.…”