心臟磁共振影像的準確分割是臨床實踐中診斷和評估許多心血管疾病的重要先決條件。深度學習最近已被廣泛應用在心臟圖像分割上。本研究以基於深度學習的方法，訓練用於分割心臟左心室、心肌及右心室的模型，然而模型會受到訓練資料的內容而有所限制，因此本研究探討模型在跨院資料上的分割表現、可行性及可靠性。我們設計了幾項實驗以提升模型的分割表現，實驗方法有半監督式學習中的偽標籤方法、迭代偽標籤、邊緣標籤以及結合前面技術的最佳化模型。對於應變分析，以最佳化模型預測並產生的左心室容積的變化曲線並計算了心臟相關指標以及相對絕對體積誤差。分析結果顯示，最終提出的最佳化模型不論是在數據及影像呈現上，皆明顯優於只訓練原始資料的基本模型。

Accurate segmentation of cardiac magnetic resonance images is a prerequisite for diagnosing and assessing many major cardiovascular diseases in clinical practice. Deep learning has recently been widely used for image segmentation. This study uses a deep learning-based method to train a model for segmenting the heart's left ventricle, myocardium, and right ventricle. However, the model’s performance is limited by the content of the training data. Therefore, this study explored the model's segmentation performance, feasibility, and reliability on cross-hospital data. We designed several experiments to improve the segmentation performance of the model, including pseudo-labeling in semi-supervised learning, iterative pseudo-labeling, and edge labeling. We combined them to produce an optimized model. For strain analysis, the left ventricular volume variation curve was generated from the left ventricular volume segmented by the optimized model, and cardiac-related metrics and relative absolute volume difference were calculated. The analysis revealed that the optimized model was considerably improved from the basic model obtained from training the original data.

1. Chen, C., et al., Deep learning for cardiac image segmentation: a review. Frontiers in Cardiovascular Medicine, 2020. 7: p. 25.
2. Hammouda, K., et al., A new framework for performing cardiac strain analysis from cine MRI imaging in mice. Scientific reports, 2020. 10(1): p. 1-15.
3. Campello, V.M., et al., Multi-Centre, Multi-Vendor and Multi-Disease Cardiac Segmentation: The M&Ms Challenge. IEEE Transactions on Medical Imaging, 2021. 40(12): p. 3543-3554.
4. Hariharan, B., et al. Hypercolumns for object segmentation and fine-grained localization. in Proceedings of the IEEE conference on computer vision and pattern recognition. 2015.
5. Ronneberger, O., P. Fischer, and T. Brox. U-net: Convolutional networks for biomedical image segmentation. in International Conference on Medical image computing and computer-assisted intervention. 2015. Springer.
6. Seyedhosseini, M., M. Sajjadi, and T. Tasdizen. Image segmentation with cascaded hierarchical models and logistic disjunctive normal networks. in Proceedings of the IEEE international conference on computer vision. 2013.
7. Çiçek, Ö., et al. 3D U-Net: learning dense volumetric segmentation from sparse annotation. in International conference on medical image computing and computer-assisted intervention. 2016. Springer.
8. Sohn, K., et al., Fixmatch: Simplifying semi-supervised learning with consistency and confidence. Advances in neural information processing systems, 2020. 33: p. 596-608.
9. Lee, D.-H. Pseudo-label: The simple and efficient semi-supervised learning method for deep neural networks. In Workshop on challenges in representation learning, ICML. 2013.
10. Rosenberg, C., M. Hebert, and H. Schneiderman, Semi-supervised self-training of object detection models. 2005.
11. Xie, Q., et al. Self-training with noisy student improves imagenet classification. in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020.
12. Ruder, S., An overview of proxy-label approaches for semi-supervised learning. 2018.
13. Bernard, O., et al., Deep Learning Techniques for Automatic MRI Cardiac Multi-Structures Segmentation and Diagnosis: Is the Problem Solved? IEEE Transactions on Medical Imaging, 2018. 37(11): p. 2514-2525.