This work is addressing the Brain Magnetic Resonance Image Synthesis for Tumor Segmentation (BraSyn) challenge which was hosted as part of the Brain Tumor Segmentation challenge (BraTS) 2023. In this challenge researchers are invited to work on synthesizing a missing magnetic resonance image sequence given other available sequences to facilitate tumor segmentation pipelines trained on complete sets of image sequences. This problem can be addressed using deep learning in the framework of paired images-to-image translation. In this work, we proposed to investigate the effectiveness of a commonly-used deep learning framework such as Pix2Pix trained under supervision of different image-quality loss functions. Our results indicate that using different loss functions significantly affects the synthesis quality. We systematically study the impact of different loss functions in the multi-sequence MR image synthesis setting of the BraSyn challenge. Furthermore, we show how image synthesis performance can be optimized by beneficially combining different learning objectives.

翻译：本研究针对脑肿瘤分割挑战赛（BraTS）2023中作为子任务举办的脑磁共振图像合成用于肿瘤分割（BraSyn）挑战赛展开。该挑战赛邀请研究者利用给定的现有序列合成缺失的磁共振图像序列，以促进基于完整图像序列训练的肿瘤分割流程。该问题可通过配对图像到图像翻译框架下的深度学习技术解决。本文拟探究在Pix2Pix等常用深度学习框架中，采用不同图像质量损失函数监督训练的合成效果。结果表明，不同损失函数的使用会显著影响合成质量。我们系统研究了BraSyn挑战赛多序列MR图像合成场景下各损失函数的影响，并进一步展示了如何通过优化组合不同学习目标来提升图像合成性能。