Multiscale Context-Cascaded Ensemble Framework (MsC²EF): Application to Breast Histopathological Image

Abstract

Microscopic analysis of breast cell and tissue is a critical step in the definitive diagnosis of breast cancer. However, it's time-consuming and fatigable for histopathologists to find the diagnostic characteristic of cell and tissue in breast histopathological image through multiple magnification scannings. Many computer-aided studies, including traditional machine learning and deep learning approaches, have been conducted to efficiently assist histopathologists in making diagnostic decision. However, precision and complexity of such approaches remain challenging. In this work, we propose and evaluate a new framework, called multiscale context-cascaded ensemble framework (MsC 2 EF), to classify breast histopathological images. The model based on MsC 2 EF exhibits a higher precision than traditional machine learning. Meantime, it is more efficient and hardware-independent compared with deep learning approaches. The MsC 2 EF consists of the input, cascade, and decision layers. The input layer comprises a feature extractor and a spatial pyramid of image to execute feature input from coarse to fine scales. Four ensemble channels are stacked in a parallel manner as the cascade layer to select and transfer contextual feature iteratively and adaptively. For the decision layer, kernel fusion-based method is integrated to perform classification of breast histopathological image by fusing four different feature spaces. Our proposed method has been evaluated on an open dataset. The experimental result shows that MsC 2 EF obtains a good classification performance (Accuracy at patch level: 0.948±0.016; accuracy at patient level: 0.981±0.016), indicating its potential application to the classification of breast histopathologist images. INDEX TERMS MsC 2 EF, breast cancer, histopathological image, ensemble learning.

A State-of-the-art Survey of Artificial Neural Networks for Whole-slide Image Analysis:from Popular Convolutional Neural Networks to Potential Visual Transformers

“…In [143], the author introduces and evaluates a frame- In [142], the author proposes a deep learning method to classify and expand PCa to achieve the purpose of classification. This method combines three independent CNNs and makes them work with patches of different sizes to obtain a large amount of contextual information.…”

A Comprehensive Review for MRF and CRF Approaches in Pathology Image Analysis

“…The characteristic of remote sensing imagery requires their analysis methods have the capacity to model multiscale object features and semantic information, thus reducing classification errors. It has shown that learning discriminative patterns from the multiscale features delivers a more robust classifier with better discriminant performance in pathology analysis [123], [124]. Maybe because it corresponds to the way that the pathologists diagnose diseases, by examining macroscopical features to microcosmic features.…”