repository
유방암 진단을 위한 SLIC 분할과 딥러닝의 융합
- Alternative Title
- Integration of SLIC segmentation and deep learning method for breast tumor classification
- Author(s)
- 박진혁
- Alternative Author(s)
- Jin Hyeok Park
- Advisor
- 선우명훈
- Department
- 일반대학원 전자공학과
- Publisher
- The Graduate School, Ajou University
- Publication Year
- 2021-08
- Language
- kor
- Abstract
- 본 논문은 의사의 시점을 고려하여 X-ray 이미지 상에서 주변 픽셀에 비해 밝은 영역의 특성을 반영하는 네트워크를 구현한 논문이다. SLIC (Simple Linear Iterative Clustering) 알고리즘을 적용하여 RoI (Region of Interest) 이미지의 지역적인 특징과 원본 X-ray 이미지의 전체적인 특징을 모두 반영한 종양 분류 모델을 고안하였다. 종양의 악성여부를 추정하는 분류 네트워크를 구현하여 RoI 미적용시 정확도 0.925, 적용시 정확도 0.96라는 유의미한 성능 차이를 보여주었다.
- Alternative Abstract
- This paper implements a deep learning network that reflects the characteristics of a region that is brighter than the surrounding pixels on an X-ray image, considering the point of view of the radiologist. We apply a Simple Linear Iterative Clustering (SLIC) algorithm to devise a tumor classification model that reflects both the local features of the Region of Interest (RoI) image and the global features of the original image. We implemented a classification network to estimate the malignancy of the tumor which showed a significant performance difference of accuracy 0.96 on application of ROI, accuracy 0.925 without ROI.
- Fulltext
- Appears in Collections:
- Graduate School of Ajou University > Department of Electronic Engineering > 3. Theses(Master)
- Files in This Item:
- There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
-
- License
- STATISTICS
- Total Visit :3,550,928
- Total Download :1,795
- Today View :13,292
