TOP > 外国特許検索 > TRAINED MODEL GENERATION PROGRAM, IMAGE GENERATION PROGRAM, TRAINED MODEL GENERATION DEVICE, IMAGE GENERATION DEVICE, TRAINED MODEL GENERATION METHOD, AND IMAGE GENERATION METHOD

TRAINED MODEL GENERATION PROGRAM, IMAGE GENERATION PROGRAM, TRAINED MODEL GENERATION DEVICE, IMAGE GENERATION DEVICE, TRAINED MODEL GENERATION METHOD, AND IMAGE GENERATION METHOD UPDATE 実績あり

外国特許コード F210010558
整理番号 (S2020-0175-N0)
掲載日 2021年10月18日
出願国 世界知的所有権機関(WIPO)
国際出願番号 2021JP006833
国際公開番号 WO 2021182103
国際出願日 令和3年2月24日(2021.2.24)
国際公開日 令和3年9月16日(2021.9.16)
優先権データ
  • 特願2020-042154 (2020.3.11) JP
発明の名称 (英語) TRAINED MODEL GENERATION PROGRAM, IMAGE GENERATION PROGRAM, TRAINED MODEL GENERATION DEVICE, IMAGE GENERATION DEVICE, TRAINED MODEL GENERATION METHOD, AND IMAGE GENERATION METHOD UPDATE 実績あり
発明の概要(英語) This trained model generation program causes a computer to implement: a learning execution function for inputting, to a machine learning device, first input image data indicating a first input image generated by a first reconstruction method using compressed sensing, and second input image data indicating a second input image generated by a second reconstruction method that is different from the first reconstruction method and is an analytical reconstruction method, and causing the machine learning device to execute machine learning and generate a trained model; and a trained model acquisition function for acquiring trained model data indicating the trained model. Input image data indicating an input image is inputted to the trained model, and a reconstructed image that has improved image quality is generated.
従来技術、競合技術の概要(英語) BACKGROUND ART
In order to reduce the exposure dose of a subject, for example, a sparse view CT (Computed Tomography) in which the number in the projection direction is reduced and the subject is scanned, and an X-ray CT device that executes X low dose for scanning the subject in a state where a current flowing through an X-ray tube is suppressed CT are used. Reconstruction methods using compressed sensing are used as the main reconstruction method to produce high quality reconstructed images when these scans are performed. The reconstruction method generates a reconstructed image using a model in which the density variation of the reconstructed image is piecewise uniform. In addition, of such reconstruction methods, a reconstruction method called Total Variation (TV: Total Variation) is frequently used in medical X ray CT devices.
An example of such a technique is the medical imaging device disclosed in Patent Document 1. The medical imaging apparatus includes an image reconstruction unit that reconstructs an image by performing iterative optimization operations of compression sensing, and a base selection unit that selects a base transformation used for optimization for each iteration. The base selecting unit selects the bases according to a predetermined base order. Alternatively, the base selecting unit selects the base by using a weighting coefficient preset as the base.
  • 出願人(英語)
  • ※2012年7月以前掲載分については米国以外のすべての指定国
  • UNIVERSITY OF TSUKUBA
  • 発明者(英語)
  • KUDO Hiroyuki
  • MORI Kazuki
国際特許分類(IPC)
参考情報 (研究プロジェクト等) Image Science Lab.
この特許について質問等ある場合は、電子メールによりご連絡ください。

PAGE TOP

close
close
close
close
close
close