Title:
MEDICAL IMAGING DEVICE AND MEDICAL IMAGE PROCESSING METHOD
Document Type and Number:
WIPO Patent Application WO/2019/026407
Kind Code:
A1
Abstract:
Provided is a medical imaging device that employs a novel image reconstruction technique in which DL is applied to medical image processing, significantly reducing the total imaging time including the time for image reconstruction and enabling the capture of a high-quality medical image in a short time. A medical image is classified into any of a plurality of predetermined kinds. One or more optimal restorers are selected on the basis of the classification results from a plurality of restorers, each thereof corresponding to each of the plurality of kinds. A reconstruction process is performed using the selected restorer. For example, the medical image is divided into a plurality of patches, and the patches are independently subjected to the reconstruction process and integrated. The restorers can be configured with a CNN.
Inventors:
OGINO MASAHIRO (JP)
NOGUCHI YOSHIMI (JP)
NOGUCHI YOSHIMI (JP)
Application Number:
PCT/JP2018/021200
Publication Date:
February 07, 2019
Filing Date:
June 01, 2018
Export Citation:
Assignee:
HITACHI LTD (JP)
International Classes:
A61B5/055; A61B5/00; A61B6/03; A61B8/14; G01N24/00; G06T1/40; G06T7/00
Domestic Patent References:
| WO2014162690A1 | 2014-10-09 |
Foreign References:
| JP2016168293A | 2016-09-23 | |||
| JP2015129987A | 2015-07-16 |
Other References:
ISHIDA, TAKAYUKI: "1. Development of image processing technique using AI", INNERVISION, vol. 32, no. 7, July 2017 (2017-07-01), pages 24, XP055575141
ROMANO, YANIV: "RAISR: Rapid and Accurate Image Super Resolution", ARXIV.ORG, ARXIV:1606.01299 V3, 2016, XP055572425, Retrieved from the Internet [retrieved on 20180814]
INAMURO, KENTO: "High- Quality MR Imaging Based on Dictionary Learning Using Training Images and Observed Signals", IEICE TECHNICAL REPORT, vol. 116, no. 395, 12 January 2017 (2017-01-12), pages 123 - 128
DONG, CHAO: "Image Super-Resolution Using Deep Convolutional Networks", ARXIV.ORG, ARXIV:1501.00092V3, 2015, XP055572436, Retrieved from the Internet [retrieved on 20180814]
ROMANO, YANIV: "RAISR: Rapid and Accurate Image Super Resolution", ARXIV.ORG, ARXIV:1606.01299 V3, 2016, XP055572425, Retrieved from the Internet
INAMURO, KENTO: "High- Quality MR Imaging Based on Dictionary Learning Using Training Images and Observed Signals", IEICE TECHNICAL REPORT, vol. 116, no. 395, 12 January 2017 (2017-01-12), pages 123 - 128
DONG, CHAO: "Image Super-Resolution Using Deep Convolutional Networks", ARXIV.ORG, ARXIV:1501.00092V3, 2015, XP055572436, Retrieved from the Internet
Attorney, Agent or Firm:
SANNOZAKA PATENT LAW FIRM (JP)
Download PDF: