Dataset Information

Automated detection of the HER2 gene amplification status in Fluorescence in situ hybridization images for the diagnostics of cancer tissues.

ABSTRACT: The human epidermal growth factor receptor 2 (HER2) gene amplification status is a crucial marker for evaluating clinical therapies of breast or gastric cancer. We propose a deep learning-based pipeline for the detection, localization and classification of interphase nuclei depending on their HER2 gene amplification state in Fluorescence in situ hybridization (FISH) images. Our pipeline combines two RetinaNet-based object localization networks which are trained (1) to detect and classify interphase nuclei into distinct classes normal, low-grade and high-grade and (2) to detect and classify FISH signals into distinct classes HER2 or centromere of chromosome 17 (CEN17). By independently classifying each nucleus twice, the two-step pipeline provides both robustness and interpretability for the automated detection of the HER2 amplification status. The accuracy of our deep learning-based pipeline is on par with that of three pathologists and a set of 57 validation images containing several hundreds of nuclei are accurately classified. The automatic pipeline is a first step towards assisting pathologists in evaluating the HER2 status of tumors using FISH images, for analyzing FISH images in retrospective studies, and for optimizing the documentation of each tumor sample by automatically annotating and reporting of the HER2 gene amplification specificities.

SUBMITTER: Zakrzewski F

PROVIDER: S-EPMC6546913 | biostudies-literature | 2019 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Automated detection of the HER2 gene amplification status in Fluorescence in situ hybridization images for the diagnostics of cancer tissues.

Zakrzewski Falk F de Back Walter W Weigert Martin M Wenke Torsten T Zeugner Silke S Mantey Robert R Sperling Christian C Friedrich Katrin K Roeder Ingo I Aust Daniela D Baretton Gustavo G Hönscheid Pia P

Scientific reports 20190603 1

The human epidermal growth factor receptor 2 (HER2) gene amplification status is a crucial marker for evaluating clinical therapies of breast or gastric cancer. We propose a deep learning-based pipeline for the detection, localization and classification of interphase nuclei depending on their HER2 gene amplification state in Fluorescence in situ hybridization (FISH) images. Our pipeline combines two RetinaNet-based object localization networks which are trained (1) to detect and classify interph ...[more]

PMID: 31160649

Similar Datasets

Project description:Abortion in cattle causes significant economic losses for cattle farmers worldwide. The diversity of abortifacients makes abortion diagnostics a complex and challenging discipline that additionally is restrained by time and economy. Microbial culture has traditionally been an important method for the identification of bacterial and mycotic abortifacients. However, it comes with the inherent bias of favoring the easy-to-culture species, e.g., those that do not require cell culture, pre-enrichment, a variety of selective growth media, or different oxygen levels for in vitro growth. Molecular methods such as polymerase chain reaction (PCR) and next-generation sequencing have been established as alternatives to traditional microbial culturing methods in several diagnostic fields including abortion diagnostics. Fluorescence in situ hybridization (FISH), a bridging microscopy technique that combines molecular accuracy with culture independence, and spatial resolution of the pathogen-lesion relation, is also gaining influence in several diagnostic fields. In this study, real-time quantitative PCR (qPCR), 16S rDNA amplicon sequencing, and FISH were applied separately and in combination in order to (i) identify potentially abortifacient bacteria without the bias of culturability, (ii) increase the diagnostic rate using combined molecular methods, (iii) investigate the presence of the difficult-to-culture zoonotic agents Coxiella burnetii, Chlamydia spp., and Leptospira spp. in bovine abortions in Denmark. Tissues from 162 aborted or stillborn bovine fetuses and placentas submitted for routine diagnostics were screened for pathogenic bacteria using 16S rDNA amplicon sequencing. Lesion association of fungal elements, as well as of selection of bacterial abortifacients, was assessed using specific FISH assays. The presence of Chlamydia spp. and chlamydia-like organisms was assessed using qPCR. The study focused on bacterial and fungal abortifacients, because Danish cattle is free from most viral abortifacients. The 16S rDNA amplicon sequencing-guided FISH approach was suitable for enhancing abortion diagnostics, i.e., the diagnostic rate for cases with tissue lesions (n = 115) was increased from 46 to 53% when compared to routine diagnostic methods. Identification of Bacillus licheniformis, Escherichia coli, and Trueperella pyogenes accounted for the majority of additional cases with an established etiology. No evidence for emerging or epizootic bacterial pathogens was found. The difficult-to-culture abortifacients were either not detected or not identified as abortifacients.

Dataset Information

Automated detection of the HER2 gene amplification status in Fluorescence in situ hybridization images for the diagnostics of cancer tissues.

Publications

Automated detection of the HER2 gene amplification status in Fluorescence in situ hybridization images for the diagnostics of cancer tissues.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets