scholarly journals Multi_Scale_Tools: A Python Library to Exploit Multi-Scale Whole Slide Images

2021 ◽  
Vol 3 ◽  
Author(s):  
Niccolò Marini ◽  
Sebastian Otálora ◽  
Damian Podareanu ◽  
Mart van Rijthoven ◽  
Jeroen van der Laak ◽  
...  
Keyword(s):  
The Cancer Genome Atlas ◽  
Cancer Tissue ◽  
Tissue Samples ◽  
Multi Scale ◽  
External Data ◽  
Single Scale ◽  
Private And Public ◽  
Processing Component ◽  
Private Datasets ◽  
Whole Slide Images

Algorithms proposed in computational pathology can allow to automatically analyze digitized tissue samples of histopathological images to help diagnosing diseases. Tissue samples are scanned at a high-resolution and usually saved as images with several magnification levels, namely whole slide images (WSIs). Convolutional neural networks (CNNs) represent the state-of-the-art computer vision methods targeting the analysis of histopathology images, aiming for detection, classification and segmentation. However, the development of CNNs that work with multi-scale images such as WSIs is still an open challenge. The image characteristics and the CNN properties impose architecture designs that are not trivial. Therefore, single scale CNN architectures are still often used. This paper presents Multi_Scale_Tools, a library aiming to facilitate exploiting the multi-scale structure of WSIs. Multi_Scale_Tools currently include four components: a pre-processing component, a scale detector, a multi-scale CNN for classification and a multi-scale CNN for segmentation of the images. The pre-processing component includes methods to extract patches at several magnification levels. The scale detector allows to identify the magnification level of images that do not contain this information, such as images from the scientific literature. The multi-scale CNNs are trained combining features and predictions that originate from different magnification levels. The components are developed using private datasets, including colon and breast cancer tissue samples. They are tested on private and public external data sources, such as The Cancer Genome Atlas (TCGA). The results of the library demonstrate its effectiveness and applicability. The scale detector accurately predicts multiple levels of image magnification and generalizes well to independent external data. The multi-scale CNNs outperform the single-magnification CNN for both classification and segmentation tasks. The code is developed in Python and it will be made publicly available upon publication. It aims to be easy to use and easy to be improved with additional functions.

scholarly journals Multi-Omic Analyses of the m5C Regulator ALYREF Reveal Its Essential Roles in Hepatocellular Carcinoma

2021 ◽  
Vol 11 ◽  
Author(s):  
Chen Xue ◽  
Yalei Zhao ◽  
Ganglei Li ◽  
Lanjuan Li
Keyword(s):  
Gene Expression ◽  
Hepatocellular Carcinoma ◽  
Immune Cell ◽  
The Cancer Genome Atlas ◽  
Cancer Tissue ◽  
Hub Genes ◽  
Specific Expression ◽  
Tissue Samples ◽  
Expression Levels ◽  
Advanced Tumor

The ALYREF protein acts as a crucial epigenetic regulator in several cancers. However, the specific expression levels and functional roles of ALYREF in cancers are largely unknown, including for hepatocellular carcinoma (HCC). In a pan-cancer tissue analysis that included HCC, we assessed the expression of ALYREF compared to normal tissues using The Cancer Genome Atlas database. Associations between ALYREF gene expression and the clinical characteristics of HCC patient samples were assessed using the UALCAN database. Kaplan-Meier plots were performed to assess HCC patient prognosis, and the TIMER database was used to explore associations between ALYREF expression and immune-cell infiltrations. The same methods were used to assess eIF4A3 expression in HCC patient samples. In addition, ALYREF- and elF4A3-related differentially expressed genes (DEGs) were determined using LinkedOmics, associated protein functionalities were predicted for positively associated DEGs, and both the TargetScan and miRDB databases were used to predict potential upstream miRNAs for control of ALYREF and eIF4A3 expression. We found that ALYREF gene expression was dysregulated in several cancers and was significantly elevated in HCC patient tissue samples and HCC cell lines. The overexpression of ALYREF was significantly related to both advanced tumor-node-metastasis stages and poor HCC prognosis. Furthermore, we found that eIF4A3 expression was significantly correlated with ALYREF expression, and that upregulated eIF4A3 was significantly associated with poor HCC patient outcomes. In the protein-protein interaction network, we identified eight hub genes based on the positively associated DEGs in common between ALYREF and eIF4A3, and the high expression levels of these hub genes were positively associated with patient clinical outcomes. In addition, we identified miR-4666a-5p and miR-6124 as potential regulators of ALYREF and eIF4A3 expression. These findings suggest that increased ALYREF expression may function as a novel biomarker for both HCC diagnosis and prognosis predictions.

scholarly journals Characterizing Immune Responses in Whole Slide Images of Cancer With Digital Pathology and Pathomics

2020 ◽  
Vol 8 (4) ◽  
pp. 133-148
Author(s):  
Rajarsi Gupta ◽  
Han Le ◽  
John Van Arnam ◽  
David Belinsky ◽  
Mahmudul Hasan ◽  
...  
Keyword(s):  
Lung Cancer ◽  
Digital Pathology ◽  
Tumor Infiltrating Lymphocytes ◽  
Screening Tests ◽  
Cancer Tissue ◽  
Tissue Samples ◽  
Wide Range ◽  
Novel Biomarkers ◽  
Infiltrating Lymphocytes ◽  
Whole Slide Images

Abstract Purpose of Review Our goal is to show how readily available Pathomics tissue analytics can be used to study tumor immune interactions in cancer. We provide a brief overview of how Pathomics complements traditional histopathologic examination of cancer tissue samples. We highlight a novel Pathomics application, Tumor-TILs, that quantitatively measures and generates maps of tumor infiltrating lymphocytes in breast, pancreatic, and lung cancer by leveraging deep learning computer vision applications to perform automated analyses of whole slide images. Recent Findings Tumor-TIL maps have been generated to analyze WSIs from thousands of cases of breast, pancreatic, and lung cancer. We report the availability of these tools in an effort to promote collaborative research and motivate future development of ensemble Pathomics applications to discover novel biomarkers and perform a wide range of correlative clinicopathologic research in cancer immunopathology and beyond. Summary Tumor immune interactions in cancer are a fascinating aspect of cancer pathobiology with particular significance due to the emergence of immunotherapy. We present simple yet powerful specialized Pathomics methods that serve as powerful clinical research tools and potential standalone clinical screening tests to predict clinical outcomes and treatment responses for precision medicine applications in immunotherapy.

A PILOT STUDY OF HER-2/NEU GENE AMPLIFICATION ASSESSED BY FLUORESCENCE IN SITU HYBRIDIZATION (FISH) IN GASTRIC CANCER

2014 ◽  
pp. 15-20
Author(s):  
Van Huy Tran ◽  
Thi Minh Thi Ha ◽  
Trung Nghia Van ◽  
Viet Nhan Nguyen ◽  
Phan Tuong Quynh Le ◽  
...  
Keyword(s):  
Gastric Cancer ◽  
In Situ Hybridization ◽  
Fluorescence In Situ Hybridization ◽  
Cancer Patients ◽  
Gene Amplification ◽  
Cancer Tissue ◽  
Tissue Samples ◽  
Her 2 ◽  
Gastric Cancer Patients

Background: HER-2/neu is a predictive biomarker for treatment of gastric cancer using trastuzumab in combination with chemotherapy. This study aimed to evaluate the status of HER-2/neu gene amplification using fluorescence in situ hybridization (FISH) in gastric cancer. Patients and methods: thirty six gastric cancer patients were assessed HER-2/neu gene amplification by FISH using PathVysionTM HER-2 DNA Probe kit (including HER-2/neu probe and CEP-17 probe) with biopsy and surgical specimens. Results: The HER-2/neu gene amplification was observed in three cases (8.3%), the HER-2/neu gene amplification rate in Lauren’s intestinal-type and diffuse-type were 11.8% and 5.2%, respectively. Conclusion: We applied successfully FISH technique with gastric cancer tissue samples. This technique could be performed as routine test in gastric cancer in order to select patients that benefit from trastuzumab in combination with chemotherapy.

scholarly journals Expression and clinical significance of miR-3615 in hepatocellular carcinoma

2021 ◽  
Vol 49 (1) ◽  
pp. 030006052098154
Author(s):  
Xin Yuan ◽  
Yize Zhang ◽  
Zujiang Yu
Keyword(s):  
Hepatocellular Carcinoma ◽  
Survival Curve ◽  
Clinicopathological Characteristics ◽  
The Cancer Genome Atlas ◽  
Prognostic Information ◽  
Ki 67 ◽  
Tissue Samples ◽  
Expression Levels ◽  
Kaplan Meier ◽  
Serum Alpha Fetoprotein

Objective To investigate the association between microRNA-3615 (miR-3615) expression and the prognosis and clinicopathological features in patients with hepatocellular carcinoma (HCC). Methods We obtained clinicopathological and genomic data and prognostic information on HCC patients from The Cancer Genome Atlas (TCGA) database. We then analyzed differences in miR-3615 expression levels between HCC and adjacent tissues using SPSS software, and examined the relationships between miR-3615 expression levels and clinicopathological characteristics. We also explored the influence of miR-3615 expression levels on the prognosis of HCC patients using Kaplan–Meier survival curve analysis. Results Based on data for 345 HCC and 50 adjacent normal tissue samples, expression levels of miR-3615 were significantly higher in HCC tissues compared with adjacent tissues. MiR-3615 expression levels in HCC patients were negatively correlated with overall survival time and positively correlated with high TNM stage, serum Ki-67 expression level, and serum alpha-fetoprotein level. There were no significant correlations between miR-3615 expression and age, sex, and pathological grade. Conclusion MiR-3615 may be a promising new biomarker and prognostic factor for HCC.

scholarly journals EP4 as a Negative Prognostic Factor in Patients with Vulvar Cancer

Cancers ◽  
2021 ◽  
Vol 13 (6) ◽  
pp. 1410
Author(s):  
Anna Buchholz ◽  
Aurelia Vattai ◽  
Sophie Fürst ◽  
Theresa Vilsmaier ◽  
Christina Kuhn ◽  
...  
Keyword(s):  
Overall Survival ◽  
Prognostic Factor ◽  
Vulvar Cancer ◽  
Cox Regression ◽  
Disease Free Survival ◽  
Common Finding ◽  
Cancer Tissue ◽  
Vulvar Carcinoma ◽  
Tissue Samples

New prognostic factors and targeted therapies are urgently needed to improve therapeutic outcomes in vulvar cancer patients and to reduce therapy related morbidity. Previous studies demonstrated the important role of prostaglandin receptors in inflammation and carcinogenesis in a variety of tumor entities. In this study, we aimed to investigate the expression of EP4 in vulvar cancer tissue and its association with clinicopathological data and its prognostic relevance on survival. Immunohistochemistry was performed on tumor specimens of 157 patients with vulvar cancer treated in the Department of Obstetrics and Gynecology, Ludwig-Maximilian-University of Munich, Germany, between 1990 and 2008. The expression of EP4 was analyzed using the well-established semiquantitative immunoreactivity score (IRS) and EP4 expression levels were correlated with clinicopathological data and patients’ survival. To specify the tumor-associated immune cells, immunofluorescence double staining was performed on tissue samples. In vitro experiments including 5-Bromo-2′-Deoxyuridine (BrdU) proliferation assay and 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromid (MTT) viability assay were conducted in order to examine the effect of EP4 antagonist L-161,982 on vulvar carcinoma cells. EP4 expression was a common finding in in the analyzed vulvar cancer tissue. EP4 expression correlated significantly with tumor size and FIGO classification and differed significantly between keratinizing vulvar carcinoma and nonkeratinizing carcinoma. Survival analysis showed a significant correlation of high EP4 expression with poorer overall survival (p = 0.001) and a trending correlation between high EP4 expression and shorter disease-free survival (p = 0.069). Cox regression revealed EP4 as an independent prognostic factor for overall survival when other factors were taken into account. We could show in vitro that EP4 antagonism attenuates both viability and proliferation of vulvar cancer cells. In order to evaluate EP4 as a prognostic marker and possible target for endocrinological therapy, more research is needed on the influence of EP4 in the tumor environment and its impact in vulvar carcinoma.

scholarly journals Comprehensive analyses of competing endogenous RNA networks reveal potential biomarkers for predicting hepatocellular carcinoma recurrence

BMC Cancer ◽  
2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Ping Yan ◽  
Zuotian Huang ◽  
Tong Mou ◽  
Yunhai Luo ◽  
Yanyao Liu ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Malignant Tumors ◽  
Expression Profiles ◽  
Clinical Information ◽  
Gene Expression Omnibus ◽  
The Cancer Genome Atlas ◽  
High Rate ◽  
Competing Endogenous Rna ◽  
Tissue Samples ◽  
Potential Biomarkers

Abstract Background Hepatocellular carcinoma (HCC) is one of the most common and deadly malignant tumors, with a high rate of recurrence worldwide. This study aimed to investigate the mechanism underlying the progression of HCC and to identify recurrence-related biomarkers. Methods We first analyzed 132 HCC patients with paired tumor and adjacent normal tissue samples from the Gene Expression Omnibus (GEO) database to identify differentially expressed genes (DEGs). The expression profiles and clinical information of 372 HCC patients from The Cancer Genome Atlas (TCGA) database were next analyzed to further validate the DEGs, construct competing endogenous RNA (ceRNA) networks and discover the prognostic genes associated with recurrence. Finally, several recurrence-related genes were evaluated in two external cohorts, consisting of fifty-two and forty-nine HCC patients, respectively. Results With the comprehensive strategies of data mining, two potential interactive ceRNA networks were constructed based on the competitive relationships of the ceRNA hypothesis. The ‘upregulated’ ceRNA network consists of 6 upregulated lncRNAs, 3 downregulated miRNAs and 5 upregulated mRNAs, and the ‘downregulated’ network includes 4 downregulated lncRNAs, 12 upregulated miRNAs and 67 downregulated mRNAs. Survival analysis of the genes in the ceRNA networks demonstrated that 20 mRNAs were significantly associated with recurrence-free survival (RFS). Based on the prognostic mRNAs, a four-gene signature (ADH4, DNASE1L3, HGFAC and MELK) was established with the least absolute shrinkage and selection operator (LASSO) algorithm to predict the RFS of HCC patients, the performance of which was evaluated by receiver operating characteristic curves. The signature was also validated in two external cohort and displayed effective discrimination and prediction for the RFS of HCC patients. Conclusions In conclusion, the present study elucidated the underlying mechanisms of tumorigenesis and progression, provided two visualized ceRNA networks and successfully identified several potential biomarkers for HCC recurrence prediction and targeted therapies.

scholarly journals Deep Learning for Automatic Subclassification of Gastric Carcinoma Using Whole-Slide Histopathology Images

Cancers ◽  
2021 ◽  
Vol 13 (15) ◽  
pp. 3811
Author(s):  
Hyun-Jong Jang ◽  
In-Hye Song ◽  
Sung-Hak Lee
Keyword(s):  
Deep Learning ◽  
Relative Proportion ◽  
The Cancer Genome Atlas ◽  
Cancer Tissue ◽  
Mucinous Tumor ◽  
Receiver Operating Characteristic Curves ◽  
Tumor Tissues ◽  
Cancer Genome Atlas ◽  
Stomach Adenocarcinoma ◽  
Tumor Types

Histomorphologic types of gastric cancer (GC) have significant prognostic values that should be considered during treatment planning. Because the thorough quantitative review of a tissue slide is a laborious task for pathologists, deep learning (DL) can be a useful tool to support pathologic workflow. In the present study, a fully automated approach was applied to distinguish differentiated/undifferentiated and non-mucinous/mucinous tumor types in GC tissue whole-slide images from The Cancer Genome Atlas (TCGA) stomach adenocarcinoma dataset (TCGA-STAD). By classifying small patches of tissue images into differentiated/undifferentiated and non-mucinous/mucinous tumor tissues, the relative proportion of GC tissue subtypes can be easily quantified. Furthermore, the distribution of different tissue subtypes can be clearly visualized. The patch-level areas under the curves for the receiver operating characteristic curves for the differentiated/undifferentiated and non-mucinous/mucinous classifiers were 0.932 and 0.979, respectively. We also validated the classifiers on our own GC datasets and confirmed that the generalizability of the classifiers is excellent. The results indicate that the DL-based tissue classifier could be a useful tool for the quantitative analysis of cancer tissue slides. By combining DL-based classifiers for various molecular and morphologic variations in tissue slides, the heterogeneity of tumor tissues can be unveiled more efficiently.

scholarly journals Two Ensemble-CNN Approaches for Colorectal Cancer Tissue Type Classification

2021 ◽  
Vol 7 (3) ◽  
pp. 51
Author(s):  
Emanuela Paladini ◽  
Edoardo Vantaggiato ◽  
Fares Bougourzi ◽  
Cosimo Distante ◽  
Abdenour Hadid ◽  
...  
Keyword(s):  
Colorectal Cancer ◽  
Deep Learning ◽  
Digital Pathology ◽  
Texture Features ◽  
Tissue Type ◽  
Cancer Tissue ◽  
Learning Methods ◽  
Feature Based ◽  
Type Classification ◽  
Whole Slide Images

In recent years, automatic tissue phenotyping has attracted increasing interest in the Digital Pathology (DP) field. For Colorectal Cancer (CRC), tissue phenotyping can diagnose the cancer and differentiate between different cancer grades. The development of Whole Slide Images (WSIs) has provided the required data for creating automatic tissue phenotyping systems. In this paper, we study different hand-crafted feature-based and deep learning methods using two popular multi-classes CRC-tissue-type databases: Kather-CRC-2016 and CRC-TP. For the hand-crafted features, we use two texture descriptors (LPQ and BSIF) and their combination. In addition, two classifiers are used (SVM and NN) to classify the texture features into distinct CRC tissue types. For the deep learning methods, we evaluate four Convolutional Neural Network (CNN) architectures (ResNet-101, ResNeXt-50, Inception-v3, and DenseNet-161). Moreover, we propose two Ensemble CNN approaches: Mean-Ensemble-CNN and NN-Ensemble-CNN. The experimental results show that the proposed approaches outperformed the hand-crafted feature-based methods, CNN architectures and the state-of-the-art methods in both databases.

scholarly journals Pan-cancer analysis of transcripts encoding novel open-reading frames (nORFs) and their potential biological functions

2021 ◽  
Vol 6 (1) ◽  
Author(s):  
Chaitanya Erady ◽  
Adam Boxall ◽  
Shraddha Puntambekar ◽  
N. Suhas Jagannathan ◽  
Ruchi Chauhan ◽  
...  
Keyword(s):  
Transcript Level ◽  
Open Reading Frames ◽  
The Cancer Genome Atlas ◽  
Small Subset ◽  
Cancer Tissue ◽  
Post Translational Modifications ◽  
Cancer Genome Atlas ◽  
Systematic Identification ◽  
Reading Frames

AbstractUncharacterized and unannotated open-reading frames, which we refer to as novel open reading frames (nORFs), may sometimes encode peptides that remain unexplored for novel therapeutic opportunities. To our knowledge, no systematic identification and characterization of transcripts encoding nORFs or their translation products in cancer, or in any other physiological process has been performed. We use our curated nORFs database (nORFs.org), together with RNA-Seq data from The Cancer Genome Atlas (TCGA) and Genotype-Expression (GTEx) consortiums, to identify transcripts containing nORFs that are expressed frequently in cancer or matched normal tissue across 22 cancer types. We show nORFs are subject to extensive dysregulation at the transcript level in cancer tissue and that a small subset of nORFs are associated with overall patient survival, suggesting that nORFs may have prognostic value. We also show that nORF products can form protein-like structures with post-translational modifications. Finally, we perform in silico screening for inhibitors against nORF-encoded proteins that are disrupted in stomach and esophageal cancer, showing that they can potentially be targeted by inhibitors. We hope this work will guide and motivate future studies that perform in-depth characterization of nORF functions in cancer and other diseases.

scholarly journals DeepRePath: Identifying the Prognostic Features of Early-Stage Lung Adenocarcinoma Using Multi-Scale Pathology Images and Deep Convolutional Neural Networks

Cancers ◽  
2021 ◽  
Vol 13 (13) ◽  
pp. 3308
Author(s):  
Won Sang Shim ◽  
Kwangil Yim ◽  
Tae-Jung Kim ◽  
Yeoun Eun Sung ◽  
Gyeongyun Lee ◽  
...  
Keyword(s):  
Lung Adenocarcinoma ◽  
Early Stage ◽  
External Validation ◽  
Area Under The Curve ◽  
The Cancer Genome Atlas ◽  
Deep Convolutional Neural Networks ◽  
Prognostic Features ◽  
Multi Scale ◽  
Model Based ◽  
Number Of Patients

The prognosis of patients with lung adenocarcinoma (LUAD), especially early-stage LUAD, is dependent on clinicopathological features. However, its predictive utility is limited. In this study, we developed and trained a DeepRePath model based on a deep convolutional neural network (CNN) using multi-scale pathology images to predict the prognosis of patients with early-stage LUAD. DeepRePath was pre-trained with 1067 hematoxylin and eosin-stained whole-slide images of LUAD from the Cancer Genome Atlas. DeepRePath was further trained and validated using two separate CNNs and multi-scale pathology images of 393 resected lung cancer specimens from patients with stage I and II LUAD. Of the 393 patients, 95 patients developed recurrence after surgical resection. The DeepRePath model showed average area under the curve (AUC) scores of 0.77 and 0.76 in cohort I and cohort II (external validation set), respectively. Owing to low performance, DeepRePath cannot be used as an automated tool in a clinical setting. When gradient-weighted class activation mapping was used, DeepRePath indicated the association between atypical nuclei, discohesive tumor cells, and tumor necrosis in pathology images showing recurrence. Despite the limitations associated with a relatively small number of patients, the DeepRePath model based on CNNs with transfer learning could predict recurrence after the curative resection of early-stage LUAD using multi-scale pathology images.

Sign in / Sign up

Export Citation Format

Share Document