Title:
METHOD AND DEVICE FOR RECONSTRUCTING PATCH-SPECIFIC THREE-DIMENSIONAL BONE MICROSTRUCTURE BASED ON DEEP LEARNING FROM CLINICAL CT SCAN DATA
Document Type and Number:
WIPO Patent Application WO/2024/075903
Kind Code:
A1
Abstract:
Various embodiments provide a method and a device for reconstructing patch-specific three-dimensional bone microstructure based on deep learning from clinical CT scan data. According to various embodiments, a computer device can divide a low-resolution skeletal system image into a plurality of low-resolution image patches, acquire each of a plurality of high-resolution image patches from the low-resolution image patches by using a pre-trained artificial neural network, assemble and post-processes the high-resolution image patches, and thus reconstruct a high-resolution bone microstructure.
Inventors:
JANG IN GWUN (KR)
CHUN BONG-JU (KR)
KOH HYUKJIN (KR)
CHUN BONG-JU (KR)
KOH HYUKJIN (KR)
Application Number:
PCT/KR2022/020533
Publication Date:
April 11, 2024
Filing Date:
December 16, 2022
Export Citation:
Assignee:
KOREA ADVANCED INST SCI & TECH (KR)
International Classes:
A61B6/00; A61B6/03; G06N3/08; G06T3/40; G06T11/00
Attorney, Agent or Firm:
YANG, Sungbo (KR)
Download PDF: