AtrialJSQnet: A New Framework for Joint Segmentation and Quantification of Left Atrium and Scars Incorporating Spatial and Shape Information

Abstract

Left atrial (LA) and atrial scar segmentation from late gadolinium enhanced magnetic resonance imaging (LGE MRI) is an important task in clinical practice. The automatic segmentation is however still challenging, due to the poor image quality, the various LA shapes, the thin wall, and the surrounding enhanced regions. Previous methods normally solved the two tasks independently and ignored the intrinsic spatial relationship between LA and scars. In this work, we develop a new framework, namely AtrialJSQnet, where LA segmentation, scar projection onto the LA surface, and scar quantification are performed simultaneously in an end-to-end style. We propose a mechanism of shape attention (SA) via an explicit surface projection, to utilize the inherent correlation between LA and LA scars. In specific, the SA scheme is embedded into a multi-task architecture to perform joint LA segmentation and scar quantification. Besides, a spatial encoding (SE) loss is introduced to incorporate continuous spatial information of the target, in order to reduce noisy patches in the predicted segmentation. We evaluated the proposed framework on 60 LGE MRIs from the MICCAI2018 LA challenge. Extensive experiments on a public dataset demonstrated the effect of the proposed AtrialJSQnet, which achieved competitive performance over the state-of-the-art. The relatedness between LA segmentation and scar quantification was explicitly explored and has shown significant performance improvements for both tasks. The code and results will be released publicly once the manuscript is accepted for publication via https://zmiclab.github.io/projects.html.

Medical Image Analysis on Left Atrial LGE MRI for Atrial Fibrillation Studies: A Review

“…In the DL-based approaches, both Chen et al (2018b) and Li et al (2020c) performed simultaneous LA and scar segmentation via multi-task learning. The simultaneous optimization scheme showed better performance than solving the two tasks independently, which ignored the intrinsic spatial relationship between LA and scars.…”

“…The simultaneous optimization scheme showed better performance than solving the two tasks independently, which ignored the intrinsic spatial relationship between LA and scars. Besides, Li et al (2020c) introduced a spatial encoding (SE) loss to incorporate continuous spatial information of the LA. Their experiments showed that the SE loss could be effective to remove noisy patches in the final predicted segmentation, and therefore evidently reduced the Hausdorff distance (HD) value.…”

“…8 shows. Even in the same dataset, one could encounter a severe data mismatch problem, resulting in poor outlier results Li et al (2020c). Therefore, large multi-center and multi-scanner datasets are needed, to validate the robustness and generalizability of proposed methods.…”

“…Instead of using the two-stage training strategy, Borra et al (2018) applied the Otsu's algorithm as a pre-processing to crop the ROI. Besides,Yu et al (2019),Wang et al (2019a) andLi et al (2020c) all employed part of the dataset from the challenge event Yu et al (2019). proposed an uncertainty-aware self-ensembling model for semi-supervised LA segmentation.…”

Recognition and Standardization of Cardiac MRI Orientation via Multi-tasking Learning and Deep Neural Networks

“…Nevertheless, recognizing the orientation of different modality CMR images and adjusting them into standard format could be as challenging as the further computing tasks. Different from other work that focuses on segmentation or classification individually [10] or combine image segmentation with quantification [6] this work proposes a DNN-based framework to solve the cardiac image segmentation and orientation recognition tasks simultaneously.…”