In this paper, the segmentation model is developed using the convolutional neural network for automatic segmentation of a right ventricle MRI image. The proposed model is trained end-to-end using an RVSC dataset that contains the right ventricle magnetic resonance images. The proposed model gives state-of-art achievement for dice metric and also for the Jaccard index. The proposed model achieves an optimal model performance of dice metric performance with 0.91 (0.10) for the training dataset and 0.88 (0.12) for the validation dataset.

中文翻译：

---

改进的卷积神经网络对磁共振图像进行右心室分割

在本文中，使用卷积神经网络开发了用于右心室MRI图像的自动分割的分割模型。使用包含右心室磁共振图像的RVSC数据集对提出的模型进行端到端训练。提出的模型为骰子度量标准以及Jaccard索引提供了最新的成果。所提出的模型实现了骰子性能的最佳模型性能，其中训练数据集为0.91（0.10），而验证数据集为0.88（0.12）。