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/

Zendy Plus

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

Zendy Tools

Zendy Open

In digital pathology, cell detection and classification are oftenprerequisites to quantify cell abundance and explore tissue spatialheterogeneity. However, these tasks are particularly challenging for multipleximmunohistochemistry (mIHC) images due to high levels of variability instaining, expression intensity, and inherent noise as a result of preprocessingartefacts. We proposed a deep learning method to detect and classify cells inmIHC whole-tumour slide images of breast cancer. Inspired by inception-v3, wedeveloped Cell COunt RegularizeD Convolutional neural Network (ConCORDe-Net)which integrates conventional dice overlap and a new cell count loss functionfor optimizing cell detection, followed by a multi-stage convolutional neuralnetwork for cell classification. In total, 20447 cells, belonging to five cellclasses were annotated by experts from 175 patches extracted from 6whole-tumour mIHC images. These patches were randomly split into training,validation and testing sets. Using ConCORDe-Net, we obtained a cell detectionF1 score of 0.873, which is the best score compared to three state of the artmethods. In particular, ConCORDe-Net excels at detecting closely located andweakly stained cells compared to other methods. Incorporating cell count lossin the objective function regularizes the network to learn weak gradientboundaries and separate weakly stained cells from background artefacts.Moreover, cell classification accuracy of 96.5% was achieved. These resultssupport that incorporating problem-specific knowledge such as cell count intodeep learning-based cell detection architectures improve the robustness of thealgorithm.

ConCORDe-Net: Cell Count Regularized Convolutional Neural Network for Cell Detection in Multiplex Immunohistochemistry Images