Analysis of Endometrium Form by Using LBP for Female Infertility
- Resource Type
- Conference
- Authors
- Mori, Kentaro; Kitaya, Kotaro; Ishikawa, Tomomoto; Hata, Yutaka
- Source
- 2018 International Conference on Machine Learning and Cybernetics (ICMLC) Machine Learning and Cybernetics (ICMLC), 2018 International Conference on. 1:75-79 Jul, 2018
- Subject
- Computing and Processing
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Pregnancy
Acoustics
Mathematical model
Image edge detection
Histograms
Motion pictures
Female infertility
Ultrasonic image
Endometrium
Local Bialy Pattern
- Language
- ISSN
- 2160-1348
In this study, we analyzed the feature of the uterine ultrasonic image in order to develop an image evaluative method to predict success of the pregnancy for female infertility patient. It is known that “Leaf pattern” is an indicator of success of pregnancy. A result using LBP found that a region of the endometrium composed by Edge. We confirmed that an angle of Edge was difference between “Homogeneous pattern and Pregnancy” and “Homogeneous pattern and No Pregnancy We evaluated a predictive rate of the pregnancy by using this angle. As an evaluation result, the overall accuracy was 69.7%, accuracy of Leaf pattern only was 60.0% and accuracy of Homogeneous pattern only 78.9%. We found that the high predictive rate was obtained for Homogeneous pattern by using this angle.