- Patent Title: Method and apparatus for classification of lesion based on learning data applying one or more augmentation methods in lesion information augmented patch of medical image
-
Application No.: US16849779Application Date: 2020-04-15
-
Publication No.: US11276490B2Publication Date: 2022-03-15
- Inventor: Helen Kim Hong , Hansang Lee , Joon Seok Lim
- Applicant: SEOUL WOMEN'S UNIVERSITY INDUSTRY-UNIVERSITY COOPERATION FOUNDATION , INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITY
- Applicant Address: KR Seoul; KR Seoul
- Assignee: SEOUL WOMEN'S UNIVERSITY INDUSTRY-UNIVERSITY COOPERATION FOUNDATION,INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITY
- Current Assignee: SEOUL WOMEN'S UNIVERSITY INDUSTRY-UNIVERSITY COOPERATION FOUNDATION,INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITY
- Current Assignee Address: KR Seoul; KR Seoul
- Agency: Studebaker & Brackett PC
- Priority: KR10-2019-0044264 20190416
- Main IPC: G06K9/00
- IPC: G06K9/00 ; G16H30/20 ; G06K9/62 ; G06T7/00 ; G06N3/08
Abstract:
A learning-based lesion classification method performed by a computer is provided. The method includes obtaining a medical image including lesion regions, extracting at least one sample region from the lesion regions within the medical image, generating LINA patch data based on the at least one sample region, and classifying the lesion regions within the medical image by performing learning based on the LINA patch data.
Public/Granted literature
Information query
