Breast cancer (BC) is the second‑leading cause of cancer‑related mortality among women worldwide. Accurate and early classification of mammographic lesions is therefore crucial for improving patient prognosis. In this study, we present a hybrid Walrus Particle Swarm Optimisation (WPS) algorithm that combines the velocity‑guided global search of Particle Swarm Optimisation (PSO) with the cooperative exploitation strategy of the Walrus Optimiser (WO). The proposed WPS is employed to tune the hyper‑parameters of a convolutional neural network enhanced with Swapping of Proficiency (CNN‑SP) for BC image classification. Numerical tests on the 29 CEC‑2017 benchmark functions demonstrate that WPS consistently reaches near‑optimal solutions, validating its exploration-exploitation balance. When applied to the CBIS‑DDSM and MIAS datasets, WPS‑CNN‑SP achieved better results than recent works. In particular, it achieved an Area Under the receiver-operating-characteristic Curve (AUC) = 98.28%, and an Accuracy = 98.99% on CBIS‑DDSM and AUC = 99.76% with Accuracy = 98.99% on MIAS dataset. These findings confirm the potential of WPS as a fast, reliable optimiser for computer‑aided BC screening systems.

Ultrasensitive multifunctional biosensor integrating ECL quenching and DPV enhancement for early classification of thyroid cancer via BRAF V600E and microRNA-221 detection.

A dynamic rescaled activation kernel network for chaotic pattern recognition and early disability risk mitigation as a biomarker in cancer classification

Breast cancer is an important topic in medical image analysis because it is a high-risk disease and the leading cause of death in women. Early detection of breast cancer improves treatment outcomes, which can be achieved by identifying it using mammography images. Computer-aided diagnostic systems detect and classify medical images of breast lesions, allowing radiologists to make accurate diagnoses. Deep learning algorithms improved the performance of these diagnoses systems. We utilized efficient deep learning approaches to propose a system that can detect breast cancer in mammograms. The proposed approach adopted relies on two main elements: improving image contrast to enhance marginal information and extracting discriminatory features sufficient to improve overall classification quality, these improvements achieved based on a new model from scratch to focus on enhancing the accuracy and reliability of breast cancer detection. The model trained on the digital database for screening mammography (DDSM) dataset and compared with different convolutional neural network (CNN) models, namely EfficientNetB1, EfficientNetB5, ResNet-50, and ResNet101. Moreover, to enhance the feature selection process, we have integrated adam optimizer in our methodology. In evaluation, the proposed method achieved 96.5% accuracy across the dataset. These results show the effectiveness of this method in identifying breast cancer through images.

Sentinel node-positive POLE-mutated endometrial cancer.

Advanced multimodal for Segmentation and Classification of breast cancer with DenseNet and Optimal Attention approach.

Classification of the breast cancer using the keypoint regions stable against the image transformations

Gated attention-enhanced vision transformer for small-sample uterine ultrasound image classification of atypical endometrial hyperplasia and endometrial cancer

Incidence rates and trends of paediatric cancer in Italy, 2008-2017.

Multi-task learning for multi-scale breast cancer ultrasound image segmentation and classification based on visual perception

A comparative study of advanced deep learning architectures for breast cancer classification on ultrasound and histological images

GR-HDUNET: GAN-refined hybrid dense U-net with ensemble for colorectal cancer classification

Precision oncology enables molecularly guided cancer therapy through multi-omics profiling, AI-driven classification, and biomarker-targeted interventions. Disulfidptosis has emerged as a promising therapeutic target, yet no ovarian cancer classification system currently incorporates this mechanism. The sequencing data of the samples in this study were obtained from TCGA and GEO databases. We analyzed 76 genes associated with disulfidptosis and performed consensus clustering based on their expression profiles to stratify the samples into two molecular subtypes. Differentially expressed genes (DEGs) were identified by comparing ovarian cancer tissues with normal samples. LASSO regression and random forest algorithms were then applied to screen marker genes that significantly influenced the clustering outcome. Ultimately, Ten disulfidptosis-related genes were ultimately selected to construct the predictive model. Single-cell sequencing was employed to characterize the tumor-specific expression patterns of key biomarkers. Digital spatial pathology analysis precisely mapped therapeutic target regions within tumor architectures. Immunohistochemical validation ultimately yielded clinically translatable biomarkers with diagnostic and therapeutic potential. Analysis of 76 disulfidptosis-related genes in ovarian cancer (OV) identified two molecular subtypes with distinct genomic profiles, tumor microenvironment characteristics, m6A regulator expression patterns, and clinical outcomes. Subgroup 1 showed copy number gains and immunosuppression, while Subgroup 2 exhibited higher tumor mutational burden (TMB) and immune activation. Subgroup 2 exhibited significantly higher immune infiltration, along with upregulated immune checkpoints. A 10-gene signature and CNN+GRU classifier robustly stratified patients. Single-cell and spatial transcriptomics confirmed epithelial-specific overexpression of key genes. This study identified a 10-gene signature related to disulfidptosis that is associated with distinct tumor microenvironment features, molecular heterogeneity, m6A modification patterns, and immune infiltration characteristics in ovarian cancer. Through multi-omics analyses-including single-cell and spatial transcriptomics-along with protein-level validation, our findings provide insights into the molecular landscape of ovarian cancer and suggest potential targets for future investigation into subtype-specific treatment strategies. All relevant code and analysis pipelines are publicly available at https://github.com/jinqy-lzu/Molecular_Subgroups_OV.git .

This research explores the importance of implementing intelligent systems based on artificial intelligence and data science for the prediction and diagnosis of pancreatic cancer, addressing a critical problem related to the high mortality rate and the difficulty of early diagnosis of this disease. It underscores its crucial role in improving survival rates and is aligned with Sustainable Development Goal 3 (SDG 3), which seeks to ensure healthy lives and promote well-being for all, as well as with SDG 9, which promotes the construction of resilient infrastructure and innovation. The overall objective of this research was to identify the best technological tools for preventive diagnosis of pancreatic cancer based on artificial intelligence and data science. This research was basic, with a qualitative approach and descriptive design, using databases such as Scopus and ScienceDirect to collect relevant information and performing documentary analysis of secondary sources. The main results reveal significant challenges, such as the selection of optimal cutoffs to balance sensitivity and specificity, the integration of clinical and genomic data, and the need for explainable models that can handle multimodal data. Notable benefits include early detection of pancreatic cancer, reduced workload for healthcare professionals, and improved diagnostic accuracy. Success stories demonstrating high levels of accuracy in pancreatic cancer classification using advanced techniques such as convolutional neural networks and deep learning are highlighted. In conclusion, the implementation of intelligent systems based on artificial intelligence is essential to improve the detection and treatment of pancreatic cancer, as they improve diagnostic accuracy and efficiency and also contribute to the creation of technological innovations with great social impact.

Early detection of breast cancer by mammography scans is crucial for improving treatment outcomes. However, low image resolutions, size, and location of lesions in dense breast tissue prove to be challenges in mammography, underscoring the importance of accurate and efficient computer-aided diagnostic systems. This article introduces a novel classification framework that utilizes histogram of oriented gradients (HOG) as a feature extractor and principal component analysis (PCA) for dimensionality reduction. Classification is implemented using the persistent homology classification algorithm (PHCA), which leverages persistent homology (PH) to capture topological properties of mammography images. The framework was evaluated on 7,632 images from the INbreast dataset with an extensive use of grid-search cross-validation to optimize preprocessing parameters. Two optimal combinations of HOG parameters and scaler were identified, with the best configuration (16 × 16 pixels per cell, 3 × 3 cells per block, and Minmax scaler) achieving strong performance. Validating on the test set, PHCA achieved an overall accuracy, precision, recall, F1-score, and specificity of 97.31%, 96.86%, 97.09%, 96.97%, and 96.86%, respectively. Clinically, the high precision (98.23%) and high recall (97.75%) for malignant cases highlight PHCA’s sensitivity in identifying malignancies, ensuring that very few malignant cases go undetected with highly trustworthy predictions. These results are shown to be competitive with existing state-of-the-art models, even exceeding in some cases, while requiring lower computational cost than deep learning-based approaches. Although the proposed method trails advanced deep models by 3–4% in some metrics, it offers a computationally efficient alternative and a potential for deployment in large-scale screening systems, demonstrating the promise of topological data analysis for breast cancer classification.

Skin cancer remains a major global health concern where early detection can significantly improve treatment outcomes. Traditional methods rely on expert evaluation, which can be prone to errors. DSSCC-Net, a deep CNN model integrated with SMOTE-Tomek oversampling, improves classification accuracy and effectively handles class imbalance in dermoscopic datasets. Trained and validated on the HAM10000, ISIC 2018 and PH2 datasets, DSSCC-Net achieved an average accuracy of 97.82% ± 0.37%, precision of 97%, recall of 97% and an AUC of 99.43%. Additional analysis using Grad-CAM and expert-labeled masks validated the model’s explainability. DSSCC-Net demonstrates state-of-the-art performance and readiness for real-world clinical integration. Current CNN-based models struggle with accurately classifying underrepresented skin lesion classes due to dataset imbalances and fail to achieve consistently high performance across diverse populations. There is a pressing need for a robust, efficient, and interpretable model to aid dermatologists in early and accurate diagnosis. This study proposes DSSCC-Net, a novel deep learning framework that integrates an optimized CNN architecture with the SMOTE-Tomek technique to address class imbalance. The model processes dermoscopic images from the HAM10000 dataset, resized to 28times28 pixels, and employs data augmentation, dropout layers, and ReLU activation to enhance feature extraction and reduce overfitting. Performance is evaluated using metrics such as accuracy, precision, recall, F1-score, and AUC, alongside Grad-CAM for interpretability. DSSCC-Net achieves a 98% classification accuracy, outperforming state-of-the-art models like VGG-16 (91.12%), ResNet-152 (89.32%), and EfficientNet-B0 (89.46%). The SMOTE-Tomek integration significantly improves minority-class detection, yielding an AUC of 99.43%. The model also demonstrates balanced precision (97%) and recall (97%), with a low loss value (0.1677), indicating strong generalization. DSSCC-Net sets a new benchmark for skin cancer classification by effectively addressing class imbalance and computational limitations. Its high interpretability, achieved through Grad-CAM, makes it a practical tool for clinical deployment. Future work includes extending this framework to other medical imaging domains and developing real-time diagnostic applications.

Objective: To develop a tool package that meets the routine statistical analysis requirements of population-based cancer registries in China based on R language, with the aim of improving data quality and efficiency, and promoting the nationwide scientific utilization of cancer registry data. Methods: The functional demands for statistical analysis of population-based cancer registry staff were collected through questionnaires or face-to-face interviews. Based on the concept of generic functions in R software's S3 object system, functions were developed by defining specific S3 classes for different data types, allowing the same function to perform diverse tasks depending on the class of input data. A stepwise development strategy was adopted to ensure logical coherence among functional modules, and all functions were systematically tested and validated in accordance with standard R package development guidelines. Results: Six categories of functions, including data reading, data manipulation, data processing, statistical calculation, visualization, and statistical reporting, were developed to support routine statistical analysis of population-based cancer registry data. Data reading functions support reading data formats required by the National Cancer Registry. Data manipulation functions empower conditional filtering of registry data and support regrouping, merging, or transforming the data based on registry attributes (such as urban/rural location) to accommodate different analytical needs. Data processing functions includes age grouping, International Classification of Diseases 10th Revision (ICD-10) classification, childhood cancer classification, and population estimation. Statistical calculation functions permit the calculation of age-standardized rates, truncated rates, cumulative rates, cumulative risks, life tables, and expansion from abridged to complete life tables. Visualization functions can generate commonly used statistical charts, including population pyramids, bar charts, and line graphs. Statistical reporting functions can integrate key indicators, charts, and narrative descriptions into comprehensive cancer registry reports. Conclusion: An R package named Canregtools was developed based on the concept of S3 generic functions. This package is free of charge, open-source, and highly efficient. It can meet the diversified needs in cancer registry data analysis, visualization, and reporting through standardized data processing workflows, thereby enhancing the quality and efficiency of routine statistical analysis in population-based cancer registries in China.

Timely and accurate diagnosis is crucial in addressing the global rise in thyroid cancer, ensuring effective treatment strategies and improved patient outcomes. We present an intelligent classification method that couples an Adaptive Convolutional Neural Network (CNN) with Cohen–Daubechies–Feauveau (CDF9/7) wavelets whose detail coefficients are modulated by an n-scroll chaotic system to enrich discriminative features. We evaluate on the public DDTI thyroid ultrasound dataset (n=1638 images; 819 malignant / 819 benign) using 5-fold cross-validation, where the proposed method attains 98.17% accuracy, 98.76% sensitivity, 97.58% specificity, 97.55% F1-score, and an AUC of 0.9912. A controlled ablation shows that adding chaotic modulation to CDF9/7 improves accuracy by +8.79 percentage points over a CDF9/7-only CNN (from 89.38 to 98.17%). To objectively position our approach, we trained state-of-the-art backbones on the same data and splits: EfficientNetV2-S (96.58% accuracy; AUC 0.987), Swin-T (96.41%; 0.986), ViT-B/16 (95.72%; 0.983), and ConvNeXt-T (96.94%; 0.987). Our method outperforms the best of these by +1.23 points in accuracy and +0.0042 in AUC, while remaining computationally efficient (28.7 ms per image; 1125 MB peak VRAM). Robustness is further supported by cross-dataset testing on TCIA (accuracy 95.82%) and transfer to an ISIC skin-lesion subset (n=28 unique images, augmented to 2048; accuracy 97.31%). Explainability analyses (Grad-CAM, SHAP, LIME) highlight clinically relevant regions. Altogether, the wavelet–chaos–CNN pipeline delivers state-of-the-art thyroid ultrasound classification with strong generalization and practical runtime characteristics suitable for clinical integration.

High-dimensional gene expression datasets pose a major challenge in cancer classification due to redundancy, noise, and the risk of overfitting. To address these issues, this study proposes a hybrid framework that integrates the Dung Beetle Optimizer (DBO) for feature selection with Support Vector Machines (SVM) for classification. DBO, a recently developed nature-inspired algorithm, effectively identifies informative and non-redundant subsets of genes by simulating dung beetles' foraging, rolling, obstacle avoidance, stealing, and breeding behaviors. The selected features are then classified using SVM with Radial Basis Function (RBF) kernels, which provide robust decision boundaries even in high-dimensional spaces. Extensive experiments were conducted on publicly available cancer-related gene expression datasets, covering binary, ternary, and quaternary classification tasks. Results show that the proposed DBO-SVM framework achieves 97.4-98.0% accuracy on binary datasets and 84-88% accuracy on multiclass datasets, with balanced Precision, Recall, and F1-scores. These findings highlight the method's ability to enhance classification performance while reducing computational cost and improving biological interpretability. The proposed hybrid model demonstrates strong potential as an efficient and reliable tool for precision medicine and biomedical data analysis.

Gene expression-based tumor classification aims to distinguish tumor types based on gene expression profiles. This task is difficult due to the high dimensionality of gene expression data and limited sample sizes. Most datasets contain tens of thousands of genes but only a small number of samples. As a result, selecting informative genes is necessary to improve classification performance and model interpretability. Many existing gene selection methods fail to produce stable and consistent results, especially when training data are limited. To address this, we propose a multi-task ensemble strategy that combines repeated sampling with joint feature selection and classification. The method generates multiple training subsets and applies multi-task logistic regression with group sparsity regularization to select a subset of genes that appears consistently across tasks. This promotes stability and reduces redundancy. The framework supports integration with standard classifiers such as logistic regression and support vector machines. It performs both gene selection and classification in a single process. We evaluate the method on simulated and real gene expression datasets. The results show that it outperforms several baseline methods in classification accuracy and the consistency of selected genes.