Title:
DEEP LEARNING BASED MEDICAL ULTRASONIC IMAGE CLASSIFICATION METHOD AND APPARATUS USING SMART DEVICE
Document Type and Number:
WIPO Patent Application WO/2019/054711
Kind Code:
A1
Abstract:
A deep learning based medical ultrasonic image classification method using a smart device according to an embodiment of the present invention comprises the steps of: receiving, through a probe, a reflected ultrasonic signal that is reflected from a target object so as to generate input data; simultaneously executing deep neural network (DNN) learning by sharing a rendering cycle performing an image restoration process with regard to the input data; and classifying a body structure of the ultrasonic image according to the execution result of the DNN learning.
Inventors:
SONG TAI-KYONG (KR)
KIM MIN (KR)
PARK MINSUK (KR)
KIM MIN (KR)
PARK MINSUK (KR)
Application Number:
PCT/KR2018/010580
Publication Date:
March 21, 2019
Filing Date:
September 11, 2018
Export Citation:
Assignee:
UNIV SOGANG RES FOUNDATION (KR)
International Classes:
G06T7/10; A61B8/08; G06K9/62; G06T1/20; G06T3/40; G06T15/00
Foreign References:
| KR20120059740A | 2012-06-11 | |||
| KR20160049897A | 2016-05-10 | |||
| KR20120031759A | 2012-04-04 | |||
| JP2016143353A | 2016-08-08 | |||
| KR20160112186A | 2016-09-28 |
Attorney, Agent or Firm:
CHUNG HYUN PATENT & LAW FIRM (KR)
Download PDF:
Previous Patent: AIR CONDITIONER FOR VEHICLE
Next Patent: BATTERY PACK HAVING COMMUNICATION TERMINAL ISOLATION FUNCTION
Next Patent: BATTERY PACK HAVING COMMUNICATION TERMINAL ISOLATION FUNCTION