[Effect of Training Data Differences on Accuracy in MR Image Generation Using Pix2pix].

Nihon Hoshasen Gijutsu Gakkai zasshi Pub Date : 2024-10-29 DOI:10.6009/jjrt.2024-1487

Masaru Tsukano, Yasushi Yamamoto, Masato Shirai, Masahiro Takamura, Kazuaki Matsuo, Yoshinori Miyahara, Yasushi Kaji

{"title":"[Effect of Training Data Differences on Accuracy in MR Image Generation Using Pix2pix].","authors":"Masaru Tsukano, Yasushi Yamamoto, Masato Shirai, Masahiro Takamura, Kazuaki Matsuo, Yoshinori Miyahara, Yasushi Kaji","doi":"10.6009/jjrt.2024-1487","DOIUrl":null,"url":null,"abstract":"Purpose: Using a magnetic resonance (MR) image generation technique with deep learning, we elucidated whether changing the training data patterns affected image generation accuracy.Methods: The pix2pix training model generated T1-weighted images from T2-weighted images or FLAIR images. Head MR images obtained at our hospital were used in this study. We prepared 300 cases for each model and four training data patterns for each model (a: 150 cases for one MR system, b: 300 cases for one MR system, c: 150 cases and augmentation data for one MR system, and d: 300 cases for two MR systems). The extension data were images of 150 cases rotated in the XY plane. The similarity between the images generated by the training and evaluation data in each group was evaluated using the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM).Results: For both MR systems, the PSNR and SSIM were higher for training dataset b than training dataset a. The PSNR and SSIM were lower for training dataset d.Conclusion: MR image generation accuracy varied among training data patterns.","PeriodicalId":74309,"journal":{"name":"Nihon Hoshasen Gijutsu Gakkai zasshi","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Nihon Hoshasen Gijutsu Gakkai zasshi","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.6009/jjrt.2024-1487","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

Purpose: Using a magnetic resonance (MR) image generation technique with deep learning, we elucidated whether changing the training data patterns affected image generation accuracy.

Methods: The pix2pix training model generated T1-weighted images from T2-weighted images or FLAIR images. Head MR images obtained at our hospital were used in this study. We prepared 300 cases for each model and four training data patterns for each model (a: 150 cases for one MR system, b: 300 cases for one MR system, c: 150 cases and augmentation data for one MR system, and d: 300 cases for two MR systems). The extension data were images of 150 cases rotated in the XY plane. The similarity between the images generated by the training and evaluation data in each group was evaluated using the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM).

Results: For both MR systems, the PSNR and SSIM were higher for training dataset b than training dataset a. The PSNR and SSIM were lower for training dataset d.

Conclusion: MR image generation accuracy varied among training data patterns.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

[训练数据差异对使用 Pix2pix 生成磁共振图像准确性的影响]。

目的：利用深度学习的磁共振（MR）图像生成技术，我们阐明了改变训练数据模式是否会影响图像生成的准确性：pix2pix训练模型从T2加权图像或FLAIR图像生成T1加权图像。本研究使用了本医院获得的头部磁共振图像。我们为每个模型准备了 300 个病例，并为每个模型准备了四种训练数据模式（a：一个 MR 系统 150 个病例；b：一个 MR 系统 300 个病例；c：一个 MR 系统 150 个病例和增强数据；d：一个 MR 系统 150 个病例和增强数据）：c：一个 MR 系统的 150 个病例和扩展数据；d：两个 MR 系统的 300 个病例）。扩展数据是在 XY 平面上旋转的 150 个病例的图像。使用峰值信噪比（PSNR）和结构相似性（SSIM）评估了每组训练数据和评估数据生成的图像之间的相似性：对于两个磁共振系统，训练数据集 b 的 PSNR 和 SSIM 均高于训练数据集 a：结论：磁共振图像生成的准确性因训练数据模式而异。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Nihon Hoshasen Gijutsu Gakkai zasshi

自引率

0.00%

发文量