Cancer Datasets Research Articles

Classification of Breast Cancer with Transfer Learning on Convolutional Neural Network Models

Breast cancer is a serious medical condition and a leading cause of death among women. Early and accurate diagnosis is crucial for improving patient outcomes. This study explores the use of Convolutional Neural Networks (CNNs) with Transfer Learning using DenseNet121 and ResNet50 models to enhance breast cancer classification via mammography. Transfer Learning enables CNN models to leverage knowledge learned from larger datasets such as ImageNet to improve performance on specific breast cancer datasets. The dataset comprised medical images with three breast variations: benign, malignant, and normal, totaling 531 data points. Data was split with a 70% training and 30% validation ratio. Two CNN models, AlexNet and ResNet50, were evaluated to compare their performance in classifying these breast cancer types. The experimental results show that AlexNet achieved a training accuracy of 98.01%, while ResNet50 achieved 64.07%. AlexNet demonstrated superior performance in identifying complex patterns in mammography images, resulting in more accurate classification of different breast cancer types. These findings highlight the potential of deep learning applications to support more precise and effective medical diagnostics for breast cancer. This research contributes significantly to the development of AI technologies in healthcare aimed at improving early detection of breast cancer. The implications of this study could expand our understanding of Transfer Learning applications in medical contexts, driving further advancements in this field to enhance patient care and prognosis

Identification of Immune Infiltration-related Molecular Features in Ovarian Cancer Patients and Experimental Validation of Immune Response Molecular Mechanisms through Integrated WGCNA, Machine Learning, and Single-cell Sequencing Analysis.

Ovarian cancer is one of the most common gynecological malignancies globally, and immunotherapy has emerged as a promising treatment strategy in recent years. However, the effectiveness of immunotherapy is often limited by immune escape mechanisms. To unravel the immune response mechanisms in ovarian cancer, this study aimed to employ integrated Weighted Gene Co-expression Network Analysis (WGCNA), machine learning, and single-- cell sequencing analysis to systematically investigate immune infiltration-related molecular features in ovarian cancer patients and experimentally validate the molecular mechanisms of the immune response. This research may provide a new theoretical foundation and treatment strategy for immune-based therapies in ovarian cancer. Relevant ovarian cancer datasets were collected from public databases. The ConsensusCluster- Plus and ggplot2 R packages were used to perform dimensionality reduction and clustering analysis of immune infiltration-related genes. Various algorithms were employed to select the best ovarian cancer prognostic model with OC consistency. The prognostic value of angiogenesis and immune-related gene expression was evaluated through Kaplan-Meier survival analysis, and the impact of immune infiltration on immune function in ovarian cancer patients was assessed. Functional pathways were identified using the Gene Set Enrichment Analysis (GSEA) method, and the infiltration abundance of immune and stromal components was inferred using the single-sample Gene Set Enrichment Analysis (ssGSEA) method. The influence of angiogenesis on the cellular level of Ovarian Cancer (OC) was explored in single- cell sequencing data, followed by in vitro cell experiments for further validation. The effect of the angiogenesis model on OC was evaluated through the above-mentioned research and experiments, aiming to investigate the mechanism of targeted therapy strategies in ovarian cancer. Immune-related data were collected from ovarian cancer patients in this study. Through WGCNA analysis, the MEturquoise module was identified, and a total of 1018 hub genes were determined. A prediction model was constructed using machine learning, with CoxBoost+StepCox selected as the best model, leading to the identification of 10 genes associated with ovarian cancer. Patients with high AIDPS had shorter survival time, and GSEA analysis revealed enrichment in immune-related pathways. Single-sample gene set enrichment analysis demonstrated increased immune cell infiltration and malignant stromal changes in the high AIDPS group. Results from in vitro cell experiments showed that silencing RPL31 inhibited the proliferation and migration of ovarian cancer cells while enhancing immune response capability. AIDPS holds significant clinical significance in Ovarian Cancer (OC) with poor prognosis observed in patients with high AIDPS. These patients exhibit more significant genomic variations, denser immune cell infiltration, and greater tolerance toward immune therapy. Importantly, inhibiting the expression of RPL31, a key component of AIDPS, can significantly suppress the proliferation, migration, and invasive properties of ovarian cancer cells, while stimulating the cytotoxicity of effector T cells and promoting immune response, thus slowing down the progression of ovarian cancer.

Automated Lung Cancer Diagnosis Applying Butterworth Filtering, Bi-Level Feature Extraction, and Sparce Convolutional Neural Network to Luna 16 CT Images.

Accurate prognosis and diagnosis are crucial for selecting and planning lung cancer treatments. As a result of the rapid development of medical imaging technology, the use of computed tomography (CT) scans in pathology is becoming standard practice. An intricate interplay of requirements and obstacles characterizes computer-assisted diagnosis, which relies on the precise and effective analysis of pathology images. In recent years, pathology image analysis tasks such as tumor region identification, prognosis prediction, tumor microenvironment characterization, and metastasis detection have witnessed the considerable potential of artificial intelligence, especially deep learning techniques. In this context, an artificial intelligence (AI)-based methodology for lung cancer diagnosis is proposed in this research work. As a first processing step, filtering using the Butterworth smooth filter algorithm was applied to the input images from the LUNA 16 lung cancer dataset to remove noise without significantly degrading the image quality. Next, we performed the bi-level feature selection step using the Chaotic Crow Search Algorithm and Random Forest (CCSA-RF) approach to select features such as diameter, margin, spiculation, lobulation, subtlety, and malignancy. Next, the Feature Extraction step was performed using the Multi-space Image Reconstruction (MIR) method with Grey Level Co-occurrence Matrix (GLCM). Next, the Lung Tumor Severity Classification (LTSC) was implemented by using the Sparse Convolutional Neural Network (SCNN) approach with a Probabilistic Neural Network (PNN). The developed method can detect benign, normal, and malignant lung cancer images using the PNN algorithm, which reduces complexity and efficiently provides classification results. Performance parameters, namely accuracy, precision, F-score, sensitivity, and specificity, were determined to evaluate the effectiveness of the implemented hybrid method and compare it with other solutions already present in the literature.

Associations between serum metabolites and female cancers: A bidirectional two-sample mendelian randomization study

Female cancers, especially breast, ovarian, cervical, and endometrial cancers, constitute a major threat to women's health worldwide. In view of the complex genetic background of cancers cannot be fully explained with current genetic information, we used a bidirectional two-sample mendelian randomization approach to explore the causal associations between serum metabolites and four major female cancers—breast, ovarian, cervical, and endometrial cancers. We analyzed the metabolites dataset from the Canadian Longitudinal Study of Aging and cancer datasets from the 10th round of the Finngen project. Replication analyses was performed with Cancer Association Consortium and Leo’s studies. Instrumental variables were analyzed using methods including the Wald ratio, inverse-variance weighted, MR-Egger, and weighted median. To ensure robustness, sensitivity analyses were performed using Cochrane’s Q, Egger’s intercept, MR-PRESSO, and leave-one-out methods. After meticulous analysis, we obtained levels of 3-hydroxyoleoylcarnitine, hexadecanedioate, tetradecanedioate, and carnitine C14 with robust causal associations with breast cancer, levels of 5alpha-androstan-3alpha,17beta-diol monosulfate (1), androstenediol (3beta,17beta) monosulfate (1), androsterone sulfate, and 5alpha-androstan-3beta,17beta-diol disulfate causal associations with endometrial cancer. The reverse analysis showed that breast, ovarian, and endometrial cancer and survival of breast and ovarian cancer were found to have causal relationships with 8, 5, 2, 6, and 3 metabolites, respectively. These insights underscore the potential roles of specific metabolites in the etiology of female cancers, providing new biomarkers for early detection, risk stratification, and disease progression monitoring. Further research could elucidate how these metabolites influence specific pathways in cancer development, offering theoretical foundations for prevention and treatment strategies.

Gene-age interaction study of breast cancer prognosis based on epigenomic data

Objective: Exploring gene-age interactions associated with breast cancer prognosis based on epigenomic data. Methods: Differential expression analysis of DNA methylation was conducted using multiple independent epigenomic datasets of breast cancer from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). The false discovery rate (FDR) method was used for multiple corrections, retaining differentially methylated sites with q-FDR≤0.05. A three-stage analytic strategy was implemented, using a multivariable Cox proportional hazards regression model to examine gene-age interactions. In the discovery phase, signals with q-FDR ≤ 0.05 were screened out using TCGA-BRCA database. In validation phaseⅠ, the interaction was validated using GSE72245 data, with criteria of P≤0.05 and consistent effect direction. In validation phaseⅡ, the signals were further validated using GSE37754 and GSE75067 data. A prognostic prediction model was constructed by incorporating clinical indicators and interaction signals. Results: The three-stage analytic strategy identified a methylation site (cg16126280EBF1), which interacted with age to jointly affect the overall survival time of patients (interaction HR= 1.001 1,95%CI:1.000 7-1.001 5,P<0.001). Stratified analysis by age showed that the effect of hypermethylation of cg16126280EBF1 was completely opposite in younger patients (HR=0.550 5, 95%CI: 0.383 8-0.789 6, P=0.001) and older patients (HR=2.166 5, 95%CI: 1.285 2-3.652 2, P=0.004). Conclusions: The DNA methylation site cg16126280EBF1 exhibits an interaction with age, jointly influencing the prognosis of breast cancer in a complex association pattern. This finding contributes new population-based evidence for the development of age-specific targeted drugs.

Survival analysis in breast cancer: evaluating ensemble learning techniques for prediction.

Breast cancer is most commonly faced with form of cancer amongst women worldwide. In spite of the fact that the breast cancer research and awareness have gained considerable momentum, there is still no one treatment due to disease heterogeneity. Survival data may be of specific interest in breast cancer studies to understand its dynamic and complex trajectories. This study copes with the most important covariates affecting the disease progression. The study utilizes the German Breast Cancer Study Group 2 (GBSG2) and the Molecular Taxonomy of Breast Cancer International Consortium dataset (METABRIC) datasets. In both datasets, interests lie in relapse of the disease and the time when the relapse happens. The three models, namely the Cox proportional hazards (PH) model, random survival forest (RSF) and conditional inference forest (Cforest) were employed to analyse the breast cancer datasets. The goal of this study is to apply these methods in prediction of breast cancer progression and compare their performances based on two different estimation methods: the bootstrap estimation and the bootstrap .632 estimation. The model performance was evaluated in concordance index (C-index) and prediction error curves (pec) for discrimination. The Cox PH model has a lower C-index and bigger prediction error compared to the RSF and the Cforest approach for both datasets. The analysis results of GBSG2 and METABRIC datasets reveal that the RSF and the Cforest algorithms provide non-parametric alternatives to Cox PH model for estimation of the survival probability of breast cancer patients.

Towards semi-supervised multi-modal rectal cancer segmentation: A large-scale dataset and a multi-teacher uncertainty-aware network

Rectal cancer is one of the most common malignant tumors of the digestive tract. Recently, deep learning has attracted significant attention in computer-aided cancerous region segmentation, offering better adjuvant treatment options for colorectal cancer patients. However, the size of existing rectal cancer datasets is insufficient, which severely limits the training of data-hungry segmentation models based on deep learning. Moreover, the existing datasets only contain image modality, while non-image modality information such as blood test reports is also important in practical clinical diagnosis. Hence, in this paper, we first collect the largest public dataset named RC4 for rectal cancer segmentation by far. RC4 contains 51,097 MRI slices from 657 patients of colorectal cancer, including MRI data and non-image data. We further propose a multi-modal multi-teacher uncertainty-aware network (M-MTUNet) for semi-supervised rectal cancer segmentation. Inspired by the human learning process, multi-teacher distillation is designed to transfer a variety of knowledge to a single student, which provides the student model with richer information. Besides, we propose a multi-modal uncertainty-aware consistency loss, which facilitates the knowledge transfer process from the two teacher networks to the student network and reduces the learning uncertainty within the student network. Extensive experiments demonstrate that M-MTUNet outperforms the current widely-used fully- and semi-supervised segmentation models.

Optimal Cut-Point Selection Methods Under Binary Classification When Subclasses Are Involved.

In practice, we often encounter binary classification problems where both main classes consist of multiple subclasses. For example, in an ovarian cancer study where biomarkers were evaluated for their accuracy of distinguishing noncancer cases from cancer cases, the noncancer class consists of healthy subjects and benign cases, while the cancer class consists of subjects at both early and late stages. This article aims to provide a large number of optimal cut-point selection methods for such setting. Furthermore, we also study confidence interval estimation of the optimal cut-points. Simulation studies are carried out to explore the performance of the proposed cut-point selection methods as well as confidence interval estimation methods. A real ovarian cancer data set is analyzed using the proposed methods.

A new competing risks model with applications to blood cancer data

Competing risks models are invaluable probabilistic models in survival analysis, a particular part of medical research. Such models deal with research problems that involve multiple potential risk factors that compete with each other to cause death (i.e., failure from a statistical perspective). Competing risks models' flexibility must be thoroughly explored to guarantee suitability for multifaceted risk scenarios (e.g., modeling data of malicious diseases where factors such as treatment response and progression are intertwined). This article considers a new competing risks model called the additive generalized linear-exponential (AGLE) competing risks model. The proposed model is expected to be more robust and superior to other well-known models when modeling real-life data. A mathematical treatment for the properties of the new model is first considered. Afterward, model parameters estimation via various estimation methods is discussed. The critical role of estimating shape parameters in understanding blood cancer's survival and failure mechanisms is highlighted. Furthermore, the bias and root mean square error of different estimation methods are examined numerically through Monte Carlo simulations. The simulation study indicated that the maximum product of spacings estimation method for the model parameters has the optimal balance of bias and variance as the sample sizes increase. Two real-life blood cancer data sets are analyzed to illustrate the application of the proposed model. The analysis outcomes support the AGLE model's robustness, with low Kolmogorov-Smirnov statistics and high p-value, affirming its superior fit for blood cancer data compared to other well-known models.

Explainable artificial intelligence in breast cancer detection and risk prediction: A systematic scoping review.

With the advances in artificial intelligence (AI), data-driven algorithms are becoming increasingly popular in the medical domain. However, due to the nonlinear and complex behavior of many of these algorithms, decision-making by such algorithms is not trustworthy for clinicians and is considered a black-box process. Hence, the scientific community has introduced explainable artificial intelligence (XAI) to remedy the problem. This systematic scoping review investigates the application of XAI in breast cancer detection and risk prediction. We conducted a comprehensive search on Scopus, IEEE Explore, PubMed, and Google Scholar (first 50 citations) using a systematic search strategy. The search spanned from January 2017 to July 2023, focusing on peer-reviewed studies implementing XAI methods in breast cancer datasets. Thirty studies met our inclusion criteria and were included in the analysis. The results revealed that SHapley Additive exPlanations (SHAP) is the top model-agnostic XAI technique in breast cancer research in terms of usage, explaining the model prediction results, diagnosis and classification of biomarkers, and prognosis and survival analysis. Additionally, the SHAP model primarily explained tree-based ensemble machine learning models. The most common reason is that SHAP is model agnostic, which makes it both popular and useful for explaining any model prediction. Additionally, it is relatively easy to implement effectively and completely suits performant models, such as tree-based models. Explainable AI improves the transparency, interpretability, fairness, and trustworthiness of AI-enabled health systems and medical devices and, ultimately, the quality of care and outcomes.

Relationship between prostate-specific antigen, alkaline phosphatase levels, and time-to-tumor shrinkage: understanding the progression of prostate cancer in a longitudinal study

BackgroundThis study delves into the complex interplay among prostate-specific antigen, alkaline phosphatase, and the temporal dynamics of tumor shrinkage in prostate cancer. By investigating the longitudinal trajectories and time-to-prostate cancer tumor shrinkage, we aim to untangle the intricate patterns of these biomarkers. This understanding is pivotal for gaining profound insights into the multifaceted aspects of prostate cancer progression. The joint model approach serves as a comprehensive framework, facilitating the elucidation of intricate interactions among these pivotal elements within the context of prostate cancer .MethodsA new joint model under a shared parameters strategy is proposed for mixed bivariate longitudinal biomarkers and event time data, for obtaining accurate estimates in the presence of missing covariate data. The primary innovation of our model resides in its effective management of covariates with missing observations. Built upon established frameworks, our joint model extends its capabilities by integrating mixed longitudinal responses and accounting for missingness in covariates, thus confronting this particular challenge. We posit that these enhancements bolster the model’s utility and dependability in real-world contexts characterized by prevalent missing data. The main objective of this research is to provide a model-based approach to get full information from prostate cancer data collected with patients’ baseline characteristics (Age\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$Age$$\\end{document}, body mass index (BMI\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$BMI$$\\end{document}), GleasonScore\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$Gleason Score$$\\end{document}, Grade\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$Grade$$\\end{document}, and Drug\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$Drug$$\\end{document}) and two longitudinal endogenous covariates (Platelets\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$Platelets$$\\end{document} and Bilirubin\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$Bilirubin$$\\end{document}).ResultsThe results reveal a clear association between prostate-specific antigen and alkaline phosphatase biomarkers in the context of time-to-prostate cancer tumor shrinkage. This underscores the interconnected dynamics of these key indicators in gauging disease progression.ConclusionsThe analysis of the prostate cancer dataset, incorporating a joint evaluation of mixed longitudinal prostate-specific antigen and alkaline phosphatase biomarkers alongside tumor status, has provided valuable insights into disease progression. The results demonstrate the effectiveness of the proposed joint model, as evidenced by accurate estimates. The shared variables associated with both longitudinal biomarkers and event times consistently deviate from zero, highlighting the robustness and reliability of the model in capturing the complex dynamics of prostate cancer progression. This approach holds promise for enhancing our understanding and predictive capabilities in the clinical assessment of prostate cancer.

Breast cancer relapse disease prediction improvements with ensemble learning approaches

Diagnosis and prognosis are especially difficult areas of medical research related to cancer due to the high incidence of breast cancer, which has surpassed all other cancers in terms of female mortality. Another factor that has a substantial influence on the quality of life of cancer patients is the fear that they may experience a relapse of their disease. The objective of the study is to give medical practitioners a more effective strategy for using ensemble learning techniques to forecast when breast cancer may recur. This research aimed to investigate the usage of deep neural networks (DNNs) and artificial neural networks (ANNs) in addition to machine learning (ML) based approaches, including bagging, averaging, and voting, to enhance the efficacy of breast cancer relapse diagnosis on two breast cancer relapse datasets. Results from the empirical study demonstrate that the proposed ensemble learning-enabled approach improves accuracies by 96.31% and 95.81%, precisions by 96.70% and 96.15%, sensitivities by 98.88% and 98.68%, specificities by 84.62% in both, F1-scores by 97.78% and 97.40%, and area under the curve (AUCs) of 0.987 and 0.978, with University Medical Centre, Institute of Oncology (UMCIO) and Wisconsin prognostic breast cancer (WPBC) datasets respectively. Consequently, these improved disease outcomes may encourage physicians to use this model to make better treatment choices.

An Optimized Framework Based on Data Exploration and Dynamic Ensemble-Based Models for Breast Cancer Prediction

Breast cancer (BC) is a major global health concern. Detecting BC at an early stage gives more treatment options and can help avoid more aggressive treatments. The use of machine learning (ML) in BC prediction offers significant potential for improving the accuracy and speed of diagnosis, personalizing treatment, and identifying high-risk patients. However, there are significant challenges associated with the use of ML, including the need for high-quality data and more flexible models with optimal parameters to achieve high efficiency. In this paper, we propose an optimized framework based on multi-stage data exploration. This framework is designed to provide a comprehensive approach to data exploration, ensuring that the data is well-prepared for ML. In addition, the framework includes dynamic ensemble-based classifiers, which combine multiple independent classifiers to improve accuracy and mitigate the risk of overfitting in conjunction with the cross-validation techniques. These classifiers are optimized using Bayesian hyperparameter tuning, which involves selecting the optimal values for the various hyperparameters of the model. This approach can significantly improve the prediction accuracy of the resulting model. The study evaluates the framework using the publicly available Wisconsin Diagnostic Breast Cancer (WDBC) dataset and compares our results with other state-of-the-art models. The experimental results show that the best result is 100% for accuracy and recall with hyperparameters of (Ensemble Method = AdaBoost, Number of learners = 322, learning rate = 0.9350, and the Maximum number of splits = 1). The highest performance has been achieved with the proposed framework compared with the other models in terms of accuracy (mean = 99.35%, best = 100%, worst = 98.7%, and Standard Deviation = 0.325). The framework can potentially improve the accuracy and efficiency of BC prediction, ultimately leading to better outcomes for patients.

MedCapsNet: A modified Densenet201 model integrated with capsule network for heel disease detection and classification

Conditions affecting the heel bone, such as heel spurs and sever's disease, pose significant challenges to patients' daily activities. While orthopedic and traumatology doctors rely on foot X-rays for diagnosis, there is a need for more AI-based detection and classification of these conditions. Therefore, this study addresses this need by proposing MedcapsNet, a novel hybrid capsule model combining modified DenseNet201 with a capsule network, designed to accurately detect and classify heel bone diseases utilizing lateral heel x-ray foot images. We conducted a comprehensive series of experiments on the proposed hybrid architecture with several datasets, including the Heel dataset, Breast BreaKHis v1, HAM10000 skin cancer dataset, and Jun Cheng Brain MRI dataset. The first experiment evaluates the proposed model for heel diseases, while the other experiments evaluate the model on a range of medical datasets to demonstrate its performance over existing studies. On the heel dataset, MedCapsNet achieves an accuracy of 96.38%, AUC of 98.35% without data augmentation, cross-validation accuracy of 95.69%, and AUC of 98.87%. The proposed model, despite employing a fixed architecture and hyperparameters, outperformed other models across four distinct datasets, including MRI, X-ray, and microscopic images with various diseases. This is notable because different types of medical image datasets typically require different architectures and hyperparameters to achieve optimal performance.

DynamicWeighted Particle Swarm Optimization - Support Vector Machine Optimization in Recursive Feature Elimination Feature Selection

Feature Selection is a crucial step in data preprocessing to enhance machine learning efficiency, reduce computational complexity, and improve classification accuracy. The main challenge in feature selection for classification is identifying the most relevant and informative subset to enhance prediction accuracy. Previous studies often resulted in suboptimal subsets, leading to poor model performance and low accuracy. This research aims to enhance classification accuracy by utilizing Recursive Feature Elimination (RFE) combined with Dynamic Weighted Particle Swarm Optimization (DWPSO) and Support Vector Machine (SVM) algorithms. The research method involves the utilization of 12 datasets from the University of California, Irvine (UCI) repository, where features are selected via RFE and applied to the DWPSO-SVM algorithm. RFE iteratively removes the weakest features, constructing a model with the most relevant features to enhance accuracy. The research findings indicate that DWPSO-SVM with RFE significantly improves classification accuracy. For example, accuracy on the Breast Cancer dataset increased from 58% to 76%, and on the Heart dataset from 80% to 97%. The highest accuracy achieved was 100% on the Iris dataset. The conclusion of these findings that RFE in DWPSO-SVM offers consistent and balanced results in True Positive Rate (TPR) and True Negative Rate (TNR), providing reliable and accurate predictions for various applications.

DriverDetector: An R package providing multiple statistical methods for cancer driver genes detection and tools for downstream analysis

Identifying driver genes in cancer is a difficult task because of the heterogeneity of cancer as well as the complex interactions among genes. As sequencing data become more readily available, there is a growing need for detecting cancer driver genes based on statistical and mathematical modeling methods. Currently, plenty of driver gene identification algorithms have been published, but they fail to achieve consistent results. In order to obtain gene sets with high confidence, we present DriverDetector, an R package providing a convenient workflow for cancer driver genes detection and downstream analysis. We develop the background mutation rate calculating module based on the distance between genes in covariate space and binomial test, followed by the driver gene selection module which integrates 11 methods, including two already recognized approaches, a de novo method, and five variants of Fisher's method which are applied to driver gene identification for the first time. Through verification on 12 TCGA datasets, each method is able to identify a set of confirmed driver genes while the number of resulting genes vary significantly across different methods. For robust driver genes detection, a voting strategy based on 10 of the statistical methods is further applied. Results show that the collective prediction based on the voting strategy demonstrates superiority in achieving the consistency of prediction while ensuring a reasonable number of predicted genes and confirmed drivers. By comparing the results of each cancer dataset, we also find that sample size has a huge impact on the number of predicted genes. For downstream analysis, DriverDetector automatically generates plenty of plots and tables to elaborate the results. We propose DriverDetector as a user-friendly tool promoting early diagnosis of cancer and the development of targeted drugs.

Infusing Weighted Average Ensemble Diversity for Advanced Breast Cancer Detection

ABSTRACTBreast cancer is a widespread health threat for women globally, often difficult to detect early due to its asymptomatic nature. As the disease advances, treatment becomes intricate and costly, ultimately resulting in elevated fatality rates. Currently, despite the widespread use of advanced machine learning (ML) and deep learning (DL) techniques, a comprehensive diagnosis of breast cancer remains elusive. Most of the existing methods primarily utilize either attention‐based deep models or models based on handcrafted features to capture and gather local details. However, both of these approaches lack the capability to offer essential local information for precise tumor detection. Additionally, the available breast cancer datasets suffer from class imbalance issue. Hence, this paper presents a novel weighted average ensemble network (WA‐ENet) designed for early‐stage breast cancer detection that leverages the ability of ensemble technique over single classifier‐based models for more robust and accurate prediction. The proposed model employs a weighted average‐based ensemble technique, combining predictions from three diverse classifiers. The optimal combination of weights is determined using the hill climbing (HC) algorithm. Moreover, the proposed model enhances overall system performance by integrating deep features and handcrafted features through the use of HOG, thereby providing precise local information. Additionally, the proposed work addresses class imbalance by incorporating borderline synthetic minority over‐sampling technique (BSMOTE). It achieves 99.65% accuracy on BUSI and 97.48% on UDIAT datasets.

Comparative Study of Statistical Techniques for Preliminary Diagnosis of Cancer Risk

The purpose of the is to elaborate models for preliminary diagnosis using statistical techniques. We compare two models for the estimation of cervical cancer risks.This article aims to compare predictive models for cervical cancer using machine learning techniques. We set up classification tables to compare the overall correct prediction rates.The data used comes from 30 cases of cervical cancer. We fitted a Logistic Regression (LR) model and trained Artificial Neural Networks (ANNs). The multicollinearity problem, usually present in modeling with numerous predictive variables, was addressed with factor analysis and Pearson Correlations.The LR model and ANN model were evaluated based on their percentage of correct classifications. The LR model achieved an accuracy of 33.33%, while the ANN model achieved an accuracy of 16.67%.Based on the percentage of correct classification, the Logistic Regression model was superior to the Neural Networks for the cervical cancer dataset. This highlights the need for further exploration of different machine learning approaches and data preprocessing techniques to improve predictive performance for cervical cancer risks.

A New Hybrid Adaptive Optimization Algorithm Based Wavelet Neural Network for Severity Level Prediction for Lung Cancer Dataset

A New Hybrid Adaptive Optimization Algorithm Based Wavelet Neural Network for Severity Level Prediction for Lung Cancer Dataset

Lightweight multimodal Cycle-Attention Transformer towards cancer diagnosis

Cancer diagnosis, prognosis and therapeutic response predictions are based on data extracted from various modalities, such as histology slides, cell graphs and molecular profiles from genomic data. In recent years, with the increasing availability of multimodal biomedical data, an increasing number of multimodal fusion strategies have been proposed to overcome the challenge of data heterogeneity and make full use of intersections and complementarities of different data sources. However, most of them have the problem of parameter redundancy and inefficiency in the process of fusion. Therefore, in this work, we present a Cycle-Attention Transformer (Cy-Atten) framework that learns a more robust joint representation for multimodal cancer survival analysis. The Cy-Atten adopts a cycling mechanism with sparse attention and yields a theoretical cost of O(N) for adding a modality, which is proven to effectively reduce the parameter redundancy. Empirically in the experiments on two cancer datasets from the TCGA database, the parameters in our fusion part are only of the order of 103 considering three modalities. In addition to being lightweight, the Cy-Atten outperforms several state-of-the-art multimodal fusion methods in survival analysis and grade classification of two cancers from the TCGA database. Our work can be reproduced at https://github.com/dlmscl/Cycle-Attention-Transformer.