Title:
UNSUPERVISED LEARNING-BASED MEDICAL DATA ANALYSIS DEVICE AND METHOD
Document Type and Number:
WIPO Patent Application WO/2023/176992
Kind Code:
A1
Abstract:
The present invention relates to an unsupervised learning-based medical data analysis device and method, in which an anomaly in medical data is detected and notified of using a generative adversarial network-based machine learning model, thereby enabling accurate and quick identification, and also enabling the reduction of both time and costs incurring from the identification.
Inventors:
KO JAE YEOUNG (KR)
BAE HYUN JIN (KR)
BAE HYUN JIN (KR)
Application Number:
PCT/KR2022/003690
Publication Date:
September 21, 2023
Filing Date:
March 16, 2022
Export Citation:
Assignee:
PROMEDIUS INC (KR)
International Classes:
G16H50/20; G06N3/08; G16H50/50; G16H50/70
Foreign References:
| KR102271740B1 | 2021-07-02 | |||
| KR20200141812A | 2020-12-21 | |||
| KR20210010769A | 2021-01-28 | |||
| KR20210064619A | 2021-06-03 |
Other References:
YONG-GEUN MOON, MIN-SEONG KWON, JUNG-HOON NOH, BYUNGJU LEE: "Anomalous Sound Detection Using Multiple Autoencoders", PROCEEDINGS OF THE KOREAN INSTITUTE OF COMMUNICATION SCIENCES CONFERENCE; FEBRUARY 9-11, 2022, KOREAN INSTITUTE OF COMMUNICATION SCIENCES CONFERENCE, KOREA, 1 February 2022 (2022-02-01) - 11 February 2022 (2022-02-11), Korea, pages 983 - 984, XP009549471
Attorney, Agent or Firm:
BLT PATENT & LAW FIRM (KR)
Download PDF: