استفاده از شبکه های پیچشی تقویتی افزایشی در طبقه بندی تصاویر بافت شناسی
Abstract
1- Introduction
2- Related work
3- Gradient boosting trees
4- Inception networks
5- Proposed approach
6- Experimental results and analysis
7- Conclusion
References
Abstract
Breast cancer is the most common cancer type diagnosed in women worldwide. While breast cancer can occur in both men and women, it is by far more prevalent in women. Researchers have developed computer-aided systems for efficient diagnosis of breast cancer from histopathological microscopic images. These systems have contributed to increased diagnosis efficiency of biopsy tissue using hematoxylin and eosin stained images. However, most computer-aided diagnosis systems have traditionally used handcrafted feature extraction methods that are both ineffective and time-consuming. In this study, we propose an approach that utilizes deep learning models with convolutional layers to extract the most useful visual features for breast cancer classification. It is shown that these deep learning models can extract better features than handcrafted feature extraction approaches. We also propose a novel boosting strategy to achieve the main goal, whereby the system is efficiently enriched by progressively combining deep learning models (weak classifiers) into a stronger classifier. Our system is used to classify hematoxylin and eosin stained breast biopsy images into two major groups (carcinomas and non-carcinomas) and four classes (normal tissues, benign lesions, in situ carcinomas and invasive carcinomas). We demonstrate applications to breast cancer histopathology images that have been considered challenging to diagnose based on conventional methodologies. Our results demonstrate that our breast cancer classifier with a boosting deep learning model significantly outperforms state-of-the-art methods.
Introduction
Breast cancer is still one of the top leading causes of death in women worldwide [27]. To diagnose a wide variety of breast cancer types properly, it is necessary to apply a medical test (commonly performed by a surgeon), followed by a microscopic analysis of breast tissue. In the first stage of this process, the doctor has to cut section biopsy materials and then stain them using hematoxylin and eosin staining. The hematoxylin solution binds deoxyribonucleic acid (DNA) and highlights nuclei, while eosin binds proteins and highlights other structures [43]. In the second stage of this analysis, pathologists evaluate tissue biopsies by visualizing highlighted regions in digitized images using microscopes. The evaluation of tissue biopsies allows the identification of early clues of tissue biopsies. However, professional pathologists must expend considerable time and effort to accomplish this task. The process of breast cancer diagnosis is not only time-consuming and expensive but also strongly depends on the prior knowledge of the pathologist and the consistency of pathologic reports. The average diagnostic accuracy of pathologists is approximately 75% [11]. Fortunately, the development of computer vision and machine learning potentially offers more reliable classification methods for the histological assessment of hematoxylin and eosin stained sections. These methods can automatically classify breast tissues into different categories with high classification rates. Thus, many researchers have developed fast and precise image analysis algorithms for breast cancer detection tasks. However, their results are still far from meeting accepted clinical requirements. For this reason, researchers have been expending most of their efforts into the development of new algorithms for histopathological image analyses [18], [41], [26]. These algorithms aim to achieve the precise classification of breast tissues as normal tissues, nonmalignant (benign) tissues, in situ carcinomas, or invasive carcinomas. In the category of benign lesions, images show changes in the normal structures of breast parenchyma that do not progress to malignancy. In situ carcinoma indicates cells that are restrained inside the mammary ductal–lobular system. Unlike in situ carcinomas, invasive carcinomas present a profile where cells spread beyond the structure of the mammary ductal–lobular system.