Stain-Robust Mitotic Figure Detection for the Mitosis Domain Generalization Challenge

Jahanifar, Mostafa; Shephard, Adam; Tajeddin, Neda Zamani; Bashir, Raja Muhammad Saad; Bilal, Mohsin; Khurram, Syed Ali; Minhas, Fayyaz; Rajpoot, Nasir M.

doi:10.1007/978-3-030-97281-3_6

Cited by 15 publications

(7 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mitosis detection has been done using the state-of-the-art “mitosis detection: fast and slow” (MDFS) method. 53 MDFS is a two-stage method where mitotic candidates are first detected using a fully convolutional neural network and then refined by a deeper CNN classifier. Several techniques have been incorporated during the training of the MDFS to make it robust against domain shift problems seen in histology images and generalize better to unseen images.…”

Section: Methodsmentioning

confidence: 99%

Cross-linking breast tumor transcriptomic states and tissue histology

Dawood,

Eastwood,

Jahanifar

et al. 2023

Cell Reports Medicine

View full text Add to dashboard Cite

Section: Methodsmentioning

confidence: 99%

Cross-linking breast tumor transcriptomic states and tissue histology

Dawood,

Eastwood,

Jahanifar

et al. 2023

Cell Reports Medicine

View full text Add to dashboard Cite

“…Mitosis detection has been done using the state-of-the-art "mitosis detection: fast and slow" (MDFS) method [52]. MDFS is a two-stage method where mitotic candidates are first detected using a fully convolutional neural network and then refined by a deeper CNN classifier.…”

Section: Estimation Of Mitotic Countsmentioning

confidence: 99%

Data-Driven Modelling of Gene Expression States in Breast Cancer and their Prediction from Routine Whole Slide Images

Dawood

Eastwood

Jahanifar

et al. 2023

Preprint

View full text Add to dashboard Cite

Identification of gene expression state of a cancer patient from routine pathology imaging and characterization of its phenotypic effects have significant clinical and therapeutic implications. However, prediction of expression of individual genes from whole slide images (WSIs) is challenging due to co-dependent or correlated expression of multiple genes. Here, we use a purely data-driven approach to first identify groups of genes with co-dependent expression and then predict their status from (WSIs) using a bespoke graph neural network. These gene groups allow us to capture the gene expression state of a patient with a small number of binary variables that are biologically meaningful and carry histopathological insights for clinically and therapeutic use cases. Prediction of gene expression state based on these gene groups allows associating histological phenotypes (cellular composition, mitotic counts, grading, etc.) with underlying gene expression patterns and opens avenues for gaining significant biological insights from routine pathology imaging directly.

show abstract

“…The mitosis detection was done using the state-of-the-art mitosis detection method called MDFS (mitosis detection: fast and slow) [58]. The MDFS method follows a two-stage approach to detect mitotic candidates.…”

Section: Estimation Of Mitotic Countsmentioning

confidence: 99%

Cancer drug sensitivity prediction from routine histology images

Dawood,

Vu,

Young

et al. 2023

Preprint

View full text Add to dashboard Cite

Drug sensitivity prediction models can aid in personalising cancer therapy, biomarker discovery, and drug design. Such models require survival data from randomized controlled trials which can be time consuming and expensive. In this proof-of-concept study, we demonstrate for the first time that deep learning can link histological patterns in whole slide images (WSIs) of Haematoxylin & Eosin (H&E) stained breast cancer sections with drug sensitivities inferred from cell lines. We employ patient-wise drug sensitivities imputed from gene expression based mapping of drug effects on cancer cell lines to train a deep learning model that predicts sensitivity to multiple drugs from WSIs. We show that it is possible to use routine WSIs to predict the drug sensitivity profile of a cancer patient for a number of approved and experimental drugs. We also show that the proposed approach can identify cellular and histological patterns associated with drug sensitivity profiles of cancer patients.

show abstract

Stain-Robust Mitotic Figure Detection for the Mitosis Domain Generalization Challenge

Cited by 15 publications

References 9 publications

Cross-linking breast tumor transcriptomic states and tissue histology

Cross-linking breast tumor transcriptomic states and tissue histology

Data-Driven Modelling of Gene Expression States in Breast Cancer and their Prediction from Routine Whole Slide Images

Cancer drug sensitivity prediction from routine histology images

Contact Info

Product

Resources

About