English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

A histopathological analysis performed by pathologists plays a key role in the diagnosis of breast cancer. A novel approach based on an image processing technique is proposed to help pathologists efficiently produce accurate diagnoses, that is composed of two modules, namely, anomaly detection with a support vector machine (ADSVM) method and a resolution adaptive network (RANet) model. The ADSVM method screens mislabeled patches to improve the training performance of the RANet model. In the RANet model, subnetworks with variable resolutions and depths are utilized to classify images according to the classification difficulty. This adaptive mechanism potentially increases the computational efficiency and prediction accuracy. The proposed RANet-ADSVM approach is evaluated using two public datasets: the BreaKHis and BACH 2018 datasets. Binary and multiclass classifications of patient and image levels at different magnification factors are conducted on the BreaKHis dataset. The best accuracies of 98.83&#x0025; and 99.14&#x0025; are obtained for the binary classification at <inline-formula> <tex-math notation="LaTeX">$200\times $ </tex-math></inline-formula> magnification at the patient and image levels, respectively. For the BACH 2018 dataset, binary and multiclass classifications on patch and image levels are performed. Experimental results show that the best accuracies for multiclass and binary classifications at the image level are 97.75&#x0025; and 99.25&#x0025;, respectively. Additionally, comparative experiments are performed and indicated that the proposed approach achieves significant improvements in both the classification accuracy and computational efficiency. Compared with similar networks (ResNet and DenseNet), the computational time is reduced by approximately 50&#x0025;.

Breast Cancer Classification From Histopathological Images Using Resolution Adaptive Network