Convolutional neural network classification of cancer cytopathology images: taking breast cancer as an example
MingXuan Xiao, Yufeng Li, Xu Yan, Min Gao, Weimin Wang
TL;DR
This paper tackles the automated classification of breast cancer cytopathology images by combining Inception-V3-based CNN feature extraction with quadtree-driven segmentation to handle high-resolution images. It introduces fixed pretrained weights, additional fully connected layers, and SoftMax classification, while aggregating per-block predictions through summation, product, or max fusion. On the BreakHis dataset, the proposed approach achieves over 0.92 accuracy across 40X–400X magnifications and demonstrates that segmentation plus fusion substantially improves both image- and patient-level performance compared to non-segmented baselines and prior methods. The work offers a practical pathway for faster, more reliable breast cancer detection and points toward richer multi-model fusion and finer-grained classification in future research $Accuracy_{image}$ and $Accuracy_{patient}$ metrics are formalized in the methodology and experiments.
Abstract
Breast cancer is a relatively common cancer among gynecological cancers. Its diagnosis often relies on the pathology of cells in the lesion. The pathological diagnosis of breast cancer not only requires professionals and time, but also sometimes involves subjective judgment. To address the challenges of dependence on pathologists expertise and the time-consuming nature of achieving accurate breast pathological image classification, this paper introduces an approach utilizing convolutional neural networks (CNNs) for the rapid categorization of pathological images, aiming to enhance the efficiency of breast pathological image detection. And the approach enables the rapid and automatic classification of pathological images into benign and malignant groups. The methodology involves utilizing a convolutional neural network (CNN) model leveraging the Inceptionv3 architecture and transfer learning algorithm for extracting features from pathological images. Utilizing a neural network with fully connected layers and employing the SoftMax function for image classification. Additionally, the concept of image partitioning is introduced to handle high-resolution images. To achieve the ultimate classification outcome, the classification probabilities of each image block are aggregated using three algorithms: summation, product, and maximum. Experimental validation was conducted on the BreaKHis public dataset, resulting in accuracy rates surpassing 0.92 across all four magnification coefficients (40X, 100X, 200X, and 400X). It demonstrates that the proposed method effectively enhances the accuracy in classifying pathological images of breast cancer.