This website requires JavaScript.

Transformer and GAN Based Super-Resolution Reconstruction Network for Medical Images

Weizhi DuHarvery Tian
Dec 2022
摘要
Because of the necessity to obtain high-quality images with minimal radiationdoses, such as in low-field magnetic resonance imaging, super-resolutionreconstruction in medical imaging has become more popular (MRI). However, dueto the complexity and high aesthetic requirements of medical imaging, imagesuper-resolution reconstruction remains a difficult challenge. In this paper,we offer a deep learning-based strategy for reconstructing medical images fromlow resolutions utilizing Transformer and Generative Adversarial Networks(T-GAN). The integrated system can extract more precise texture information andfocus more on important locations through global image matching aftersuccessfully inserting Transformer into the generative adversarial network forpicture reconstruction. Furthermore, we weighted the combination of contentloss, adversarial loss, and adversarial feature loss as the final multi-taskloss function during the training of our proposed model T-GAN. In comparison toestablished measures like PSNR and SSIM, our suggested T-GAN achieves optimalperformance and recovers more texture features in super-resolutionreconstruction of MRI scanned images of the knees and belly.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答