의생명 이미지 분류 기법의 최신 동향
Recent Trend of Biomedical Image Classification Methods
0
조회
0
다운로드
3
인용
0
참고문헌
키워드
초록
의학, 생물학 관련 영상 장비들과 기계 학습 기술이 발전하면서 의생명(biomedical) 이미지 분석은 생물학자와 기계 학습 연구자 모두에게 인기 있는 주제가 되었다. 의생명 이미지 분석에서는 분류(classification), 분할(segmentation), 이미지 정렬(registration) 등에 관한 여러 가지 기법을 연구하지만, 이 논문에서는 그 중에서도 의생명 이미지의 최신 분류 기법을 소개하였다. 바이오 분야에서 분석하는 데이터는 대부분 현미경 이미지로 이루어져 있고 의료 분야에서는 CT, MRI 등 영상 의학 데이터를 사용하므로, 분석하는 데이터의 종류에 따라 바이오 분야와 의료 분야로 나누어서 서술하였다. 또한 최근에는 이미지 분석에서 딥러닝을 많이 사용하고 있기 때문에, 의생명 이미지 분류 문제에 딥러닝을 적용한 사례들을 별도로 소개함으로써 이미지 처리 분야의 최신 아이디어를 생물학, 의학 영역에 적용하는 방법의 가능성을 살펴보았다.
Abstract
As biomedical imaging equipment and machine learning algorithms are improved, biomedical image analysis became a popular topic for both biologists and machine learning researchers. Biomedical image analysis includes various topics such as classification, segmentation, and registration. All of them are being actively studied, and there are a lot of remarkable papers on these topics. In this paper, we focus on recent trend of biomedical image classification. Because researchers use microscopy images for biological image analysis and use radiological data such as CT, MRI for medical image analysis, we explain classification methods used in several researches for biological image and medical image separately, depending on the type of images to be analyzed. In addition to traditional methods based on feature descriptor, we also introduce methods that apply deep learning in biomedical image classification, since deep learning is recently used in many researches for image processing. We found that deep learning based models show great performance in biomedical domain, and state-of-the-art idea of image processing and computer vision has potential to be applied to biomedical problems. Finally we suggest future works for better biomedical image classifier, based on the idea that are recently studied in computer vision, but with few papers in biomedical domain.
참고문헌 (14)
- G. Myers, "Why bioimage informatics matters", Nature Methods, 9(7), pp.659~ (2012)
- Q. Zhong, "Unsupervised modeling of cell morphology dynamics for time-lapse microscopy", Nature Methods, 9(7), pp.711~713 (2012)
- Y. Song, "Bioimage classification with subcategory discriminant transform of high dimensional visual descriptors", BMC Bioinformatics, 17(1), pp.465~479 (2016)
- L. Shamir, "IICBU 2008: a proposed benchmark suite for biological image analysis", Medical & Biological Engineering & Computing, 46(9), pp.943~947 (2008)
- K. Simonyan, "Very deep convolutional networks for large-scale image recognition", arXiv preprint arXiv:1409.1556, ~ (2014)
- H. J. W. L. Aerts, "Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach", Nature Communications, 5 (2014)
- M. Zhou, "Knowledge discovery and predictive modeling from brain tumor MRIs", pp.2015
- O. Russakovsky, "Imagenet large scale visual recognition challenge", International Journal of Computer Vision, 115(3), pp.211~252 (2015)
- A. Krizhevsky, "Imagenet classification with deep convolutional neural networks", Advances in Neural Information Processing Systems (2012)
- K. He, "Deep residual learning for image recognition", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2016)
- Y. LeCun, "Deep learning", Nature, 521(7553), pp.436~444 (2015)
- X. Zhu, "Deep convolutional neural network for survival analysis with pathological images", 2016 IEEE International Conference on Bioinformatics and Biomedicine (2016)
- D. R. Cox, "Regression models and life-tables", Journal of the Royal Statistical Society. Series B (Methodological), 34(2), pp.187~220 (1972)
- J. Katzman, "Deep survival: A deep cox proportional hazards network", arXiv preprint arXiv:1606.00931, ~ (2016)