Deep Learning for Virtual Histological Staining of Bright-Field Microscopic Images of Unlabeled Carotid Artery Tissue
- PDF / 4,169,017 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 43 Downloads / 183 Views
RESEARCH ARTICLE
Deep Learning for Virtual Histological Staining of Bright-Field Microscopic Images of Unlabeled Carotid Artery Tissue Dan Li,1,2 Hui Hui,2,3 Yingqian Zhang,4 Wei Tong,4 Feng Tian,4 Xin Yang,2 Jie Liu,1 Yundai Chen,4 Jie Tian 2,3,5 1
Department of Biomedical Engineering, School of Computer and Information Technology, Beijing Jiaotong University, Beijing, 100044, China 2 CAS Key Laboratory of Molecular Imaging, Institute of Automation, Chinese Academy of Sciences, Institute of Automation, Beijing, 100190, China 3 University of Chinese Academy of Sciences, Beijing, 100190, China 4 Department of Cardiology, Chinese PLA General Hospital, Beijing, 100853, China 5 Beijing Advanced Innovation Center for Big Data-Based Precision Medicine, School of Medicine, Beihang University, Beijing, 100083, China
Abstract Purpose: Histological analysis of artery tissue samples is a widely used method for diagnosis and quantification of cardiovascular diseases. However, the variable and labor-intensive tissue staining procedures hinder efficient and informative histological image analysis. Procedures: In this study, we developed a deep learning-based method to transfer bright-field microscopic images of unlabeled tissue sections into equivalent bright-field images of histologically stained versions of the same samples. We trained a convolutional neural network to build maps between the unstained images and histologically stained images using a conditional generative adversarial network model. Results: The results of a blind evaluation by board-certified pathologists illustrate that the virtual staining and standard histological staining images of rat carotid artery tissue sections and those involving different types of stains showed no major differences. Quantification of virtual and histological H&E staining in carotid artery tissue sections showed that the relative errors of intima thickness, intima area, and media area were lower than 1.6 %, 5.6 %, and 12.7 %, respectively. The training time of deep learning network was 12.857 h with 1800 training patches and 200 epoches. Conclusions: This virtual staining method significantly mitigates the typically laborious and timeconsuming histological staining procedures and could be augmented with other label-free microscopic imaging modalities. Key words: Virtual histological staining, Conditional generative adversarial network, Blind evaluation, Bright-field microscopic imaging
Dan Li, Hui Hui and Yingqian Zhang contributed equally to this work. Electronic supplementary material The online version of this article (https:// doi.org/10.1007/s11307-020-01508-6) contains supplementary material, which is available to authorized users. Correspondence to: Jie Liu; e-mail: [email protected], Yundai Chen; email: [email protected], Jie Tian; e-mail: [email protected]
Introduction Coronary artery disease (CAD) is the leading cause of mortality globally. Fundamental studies regarding CAD
Li D. et al.: Deep Learning for Virtual Histological Images
pathophysiological mechanisms
Data Loading...
