scholarly journals A Means of Assessing Deep Learning-Based Detection of ICOS Protein Expression in Colon Cancer

Cancers ◽  
2021 ◽  
Vol 13 (15) ◽  
pp. 3825
Author(s):  
Md Mostafa Kamal Sarker ◽  
Yasmine Makhlouf ◽  
Stephanie G. Craig ◽  
Matthew P. Humphries ◽  
Maurice Loughrey ◽  
...  
Keyword(s):  
Deep Learning ◽  
T Cell ◽  
Protein Expression ◽  
Open Source ◽  
Open Source Software ◽  
Digital Pathology ◽  
Response To Therapy ◽  
Cell Segmentation ◽  
Nuclear Staining ◽  
Annotation Process

Biomarkers identify patient response to therapy. The potential immune-checkpoint biomarker, Inducible T-cell COStimulator (ICOS), expressed on regulating T-cell activation and involved in adaptive immune responses, is of great interest. We have previously shown that open-source software for digital pathology image analysis can be used to detect and quantify ICOS using cell detection algorithms based on traditional image processing techniques. Currently, artificial intelligence (AI) based on deep learning methods is significantly impacting the domain of digital pathology, including the quantification of biomarkers. In this study, we propose a general AI-based workflow for applying deep learning to the problem of cell segmentation/detection in IHC slides as a basis for quantifying nuclear staining biomarkers, such as ICOS. It consists of two main parts: a simplified but robust annotation process, and cell segmentation/detection models. This results in an optimised annotation process with a new user-friendly tool that can interact with1 other open-source software and assists pathologists and scientists in creating and exporting data for deep learning. We present a set of architectures for cell-based segmentation/detection to quantify and analyse the trade-offs between them, proving to be more accurate and less time consuming than traditional methods. This approach can identify the best tool to deliver the prognostic significance of ICOS protein expression.

scholarly journals A Novel Patient-specific Protein Expression-based Algorithmic Model for Immune Checkpoint Inhibition Therapy Response Prediction in a Pan-cancer Cohort

2019 ◽  
Author(s):  
Oluwaseun Adebayo Bamodu ◽  
Wei-Hong Cheng ◽  
Oluleke Bamodu ◽  
Wei-Hwa Lee ◽  
Kang-Yun Lee ◽  
...  
Keyword(s):  
T Cell ◽  
Protein Expression ◽  
Immune Checkpoint ◽  
Drug Response ◽  
Response Prediction ◽  
Response To Therapy ◽  
Patient Specific ◽  
Beta Catenin ◽  
Statistical Regression ◽  
Objective Evidence

Background: Accurate prediction of patients’ response to therapy is clinically indispensable, howbeit challenging. With increased understanding of the human genome and malignancies, there is the renaissance of in silico pharmacogenomics with renewed interest in drug response predictability based on gene-drug interaction. Objective: Evidence-based transcript-proteome profiling is essential for synthesizing clinically applicable algorithms for predicting response to anticancer therapy, including immune checkpoint blockade (ICBT); thus, saving physicians’ time, reducing polypharmacy, and curtailing unnecessary treatment expense. In this study, we tested and validated the hypothesis that a selected proteomic signature in ICBT-naïve patients is sufficient for the prediction of response to ICBT. Methods: Using a multimodal approach consisting of computational pharmacogenomics, transcript-proteome analytics, mathematical modeling, and machine learning systems; we delineated therapy-sensitivity and stratified patients into graduated response groups based on their proteomic profile. Protein expression levels in our cohort tissue specimens were evaluated based on T cell- and non-T cell- inflamed phenotypes by immunohistochemistry. Results: We established β-catenin, PDL1, CD3 and CD8 expression-based ICBT response model. Statistical regression models validated the predictive association between our predefined algorithms and therapeutic outcome. Interestingly, our 4-gene prediction classifier was constitutively independent of tumor tissue origin, correctly stratified patients into high-, low-, and non- responders pre-treatment, with high prediction accuracy, and exhibited good association with patients’ performance status and prognosis (p < 0.01). Conclusion: Our findings demonstrate the possibility of accurate proteomics based ICBT response prediction and provide a putative basis for drug response prediction based on selective proteome profile in untreated cancer patients.

scholarly journals COMPARATIVE FEATURES OF OPEN SOURCE SOFTWARE PRODUCTS FOR THE DEVELOPMENT OF AN AUTOMATED BREAST CANCER DIAGNOSTIC PROGRAM

2019 ◽  
Vol 7 (4) ◽  
pp. 377-385
Author(s):  
V. Kovalev ◽  
Y. Diachenko ◽  
V. Malyshev ◽  
S. Rjabceva ◽  
O. Kolomiets ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Open Source ◽  
Cancer Diagnosis ◽  
Open Source Software ◽  
Imaging Techniques ◽  
Digital Pathology ◽  
Third Party ◽  
Histopathological Diagnosis ◽  
Histological Verification ◽  
Whole Slide Images

Breast cancer is one of the most common cancer diseases in the world among women. The reliability of histological verification of breast cancer depends on pathologist’s experience, knowledge, his willingness to self-improve and study specialized literature. Digital pathology is also widely used for educational purposes, in telepathology, teleconsultation and research projects. Recently developed Whole Slide Image (WSI) system opens great opportunities in the histopathological diagnosis quality improvement. Digital whole-slide images provide the effective use of morphometry and various imaging techniques to assist pathologists in quantitative and qualitative evaluation of histopathological preparations. The development of software for morphological diagnosis is important for improving the quality of histological verification of diagnosis in oncopathology. The purpose of this work is to find and benchmark existing open-source software for the whole-slide histological images processing. Choosing an open source program is an important step in developing an automated breast cancer diagnosis program. The result is a detailed study of open-source software: ASAP, Orbit, Cytomine and QuPath. Their features and methods of image processing were analyzed. QuPath software has the best characteristics for extending it with an automated module for the cancer diagnosis. QuPath combines a user-friendly, easy-to-use interface, customizable functionality, and moderate computing power requirements. Besides, QuPath works with whole-slide images with immunohistochemical markers; features implemented in this software allow making a morphometric analysis. QuPath saves time for a graphical user interface development and provides a scalable system to add new key features. QuPath supports third-party MATLAB and Python extensions.

Expression of mEGFR, pEGFRThr654, and pPKN1Thr774 to predict response to chemoradiation and prognosis in patients with cervical cancer.

2013 ◽  
Vol 31 (15_suppl) ◽  
pp. e16531-e16531
Author(s):  
Jin Yi Lang ◽  
Jianming Huang ◽  
Xin Lai
Keyword(s):  
Cervical Cancer ◽  
Protein Expression ◽  
Cancer Patients ◽  
Concurrent Chemoradiotherapy ◽  
Nuclear Translocation ◽  
Poor Response ◽  
Response To Therapy ◽  
Nuclear Staining ◽  
Poor Overall Survival ◽  
Cytoplasmic Staining

e16531 Background: The aim of this study was to determine the relation between proteins involved in phosphorylation of EGFRThr654 and response to chemoradiation and survival in a well-documented series of cervical cancer patients. Methods: Pre-treatment tissue samples of 90 consecutive FIGO stage IIA-IIIB cervical cancer patients treated with concurrent chemoradiotherapy between January 2007 and December 2009 were collected. Clinicopathologic and follow-up data were collected by a retrospective chart review. Protein expression of membranous EGFR (mEGFR), pEGFRThr654 and pPKN1Thr774 were examined by immunohistochemistry on FFPE tumor specimens. The correlations between protein expression and clinicopathological factors and outcomes were analyzed. Results: mEGFR staining was present in 67.78%, pEGFRThr654 in 91.11%, pPKN1Thr774 in 56.67% of tumors. pEGFRThr654 staining was positively correlated to pPKN1Thr774 (p = 0.04) and to mEGFR staining (p = 0.015). In multivariate analysis, nuclear staining of pEGFRThr654 [hazard ratio(HR)= 4.833; 95%CI=1.959-11.922, P=0.001)] and cytoplasmic staining of pPKN1Thr774 (HR=3.095; 95%CI=1.019-9.406; P=0.046) were prognostic factors for poor overall survival (OS) and also were independent predictors of poor response to (chemo)radiation for progress-free survival (PFS) (HR=2.921, 95%CI=1.360-6.274, P=0.006; HR=2.963, 95% CI=1.187-7.394, P=0.020), and mEGFR staining was an independent prognostic factor for poor PFS (HR=5.934, 95% CI=1.378-25.555, p=0.017). The high expression of pEGFRthr654 is related to poor local control rate (p = 0.000) and to poor short-term efficacy (p = 0.009). Conclusions: mEGFR and pEGFRThr654 immunostaining are frequently observed and independently associated with poor response to therapy and poor PFS and OS in cervical cancer patients primarily treated by concurrent chemoradiotherapy. Our data presents the pEGFRThr654 nuclear translocation as a promising target in Integrated therapies of chemoradiotherapy combined with targeting inhibition of pEGFRThr654 and its nuclear import for cervical cancer.

scholarly journals FastPathology: An open-source platform for deep learning-based research and decision support in digital pathology

IEEE Access ◽  
2021 ◽  
pp. 1-1
Author(s):  
Andre Pedersen ◽  
Marit Valla ◽  
Anna M. Bofin ◽  
Javier Perez De Frutos ◽  
Ingerid Reinertsen ◽  
...  
Keyword(s):  
Deep Learning ◽  
Decision Support ◽  
Open Source ◽  
Digital Pathology

scholarly journals QuPath: Open source software for digital pathology image analysis

2017 ◽  
Vol 7 (1) ◽  
Author(s):  
Peter Bankhead ◽  
Maurice B. Loughrey ◽  
José A. Fernández ◽  
Yvonne Dombrowski ◽  
Darragh G. McArt ◽  
...  
Keyword(s):  
Image Analysis ◽  
Open Source ◽  
Open Source Software ◽  
Digital Pathology ◽  
Pathology Image

scholarly journals IoT-based group size prediction and recommendation system using machine learning and deep learning techniques

2021 ◽  
Vol 3 (2) ◽  
Author(s):  
Deepti Chopra ◽  
Arvinder Kaur
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Open Source ◽  
Group Size ◽  
Open Source Software ◽  
Multilayer Perceptron ◽  
Recommendation System ◽  
Regression Trees ◽  
Learning Models ◽  
Information Foraging

AbstractIn an open source software development environment, it is hard to decide the number of group members required for resolving software issues. Developers generally reply to issues based totally on their domain knowledge and interest, and there are no predetermined groups. The developers openly collaborate on resolving the issues based on many factors, such as their interest, domain expertise, and availability. This study compares eight different algorithms employing machine learning and deep learning, namely—Convolutional Neural Network, Multilayer Perceptron, Classification and Regression Trees, Generalized Linear Model, Bayesian Additive Regression Trees, Gaussian Process, Random Forest and Conditional Inference Tree for predicting group size in five open source software projects developed and managed using an open source development framework GitHub. The social information foraging model has also been extended to predict group size in software issues, and its results compared to those obtained using machine learning and deep learning algorithms. The prediction results suggest that deep learning and machine learning models predict better than the extended social information foraging model, while the best-ranked model is a deep multilayer perceptron((R.M.S.E. sequelize—1.21, opencv—1.17, bitcoin—1.05, aseprite—1.01, electron—1.16). Also it was observed that issue labels helped improve the prediction performance of the machine learning and deep learning models. The prediction results of these models have been used to build an Issue Group Recommendation System as an Internet of Things application that recommends and alerts additional developers to help resolve an open issue.

Sign in / Sign up

Export Citation Format

Share Document