Automated lung cancer classification using intensity-driven RoI selection and transfer learning.

Multi-class lung cancer detection and classification from CT scans using CNN transformer

Deep learning driven early detection of lung cancer from CT images using transfer learning approaches

The proposed system demonstrates potential as an assistive tool for automated lung cancer screening, warranting further validation on larger and multi-institutional datasets before clinical application.

An Integrated Computational Framework for Lung Cancer Detection: Combining Deep Learning, Interactive Visualization, and Diagnostic Reporting from CT Imaging

Lung cancer remains the leading cause of cancer-related mortality worldwide, necessitating early and accurate diagnostic solutions. This paper presents an integrated computational framework for lung cancer detection from CT imaging, combining deep learning models with interactive visualization and automated diagnostic reporting. The proposed system leverages two state-of-the-art architectures—EfficientNet with CBAM attention enhancement and Vision Transformer including Swin Transformer variants—to achieve robust classification performance. EfficientNet provides parameter-efficient feature extraction through its compound scaling strategy, while Vision Transformers capture global contextual relationships via self-attention mechanisms, addressing the inherent limitations of CNNs in modeling long-range dependencies in medical images. The framework integrates a hybrid feature fusion approach, interactive visualization modules for clinician interpretability, and automated diagnostic report generation. Experimental evaluation on benchmark datasets demonstrates superior performance, with the optimized CBAM-EfficientNet achieving 99.81% accuracy and the ViT-based fusion approach achieving 99.28% accuracy. The system's interactive visualization capabilities, including Grad-CAM attention maps, enhance clinical interpretability and trustworthiness. This research contributes a comprehensive end-to-end solution bridging computational innovation with clinical practice for improved lung cancer diagnosis.

Capsule Network-Based Framework for Lung Cancer Classification from CT Images

Lung cancer remains one of the most critical global health challenges due to its high mortality rate and the complexity of early diagnosis. Computed Tomography (CT) imaging has become an essential modality for identifying pulmonary abnormalities; however, accurate classification of benign and malignant lung lesions remains challenging because of variations in nodule size, shape, texture, and anatomical location. Traditional computer-aided diagnosis approaches have demonstrated effectiveness but often struggle with maintaining spatial relationships between extracted features. This research presents a Capsule Network-Based Framework for Lung Cancer Classification from CT Images, integrating capsule representation learning and dynamic routing mechanisms to improve discriminative feature modeling. The proposed framework focuses on preserving hierarchical spatial information while distinguishing malignant characteristics from benign patterns. The methodology combines CT image preprocessing, lung region segmentation, feature encoding through capsule layers, and dynamic routing-based feature aggregation for classification. Existing studies on deep learning-based lung nodule analysis demonstrate the increasing importance of automated diagnostic systems, while limitations in conventional convolutional approaches motivate the exploration of capsule-based architectures (Gu et al., 2021; Shrestha and Mahmood, 2019). The framework provides a theoretically grounded approach for enhancing medical image interpretation by improving feature robustness and reducing dependence on manual analysis. The research contributes a structured deep learning model for intelligent lung cancer assessment and highlights opportunities for future clinical decision-support applications.

Deep Learning-Based Lung Cancer Segmentation and Volumetric Analysis Using CT Imaging: A Comprehensive Survey.

Although deep learning-based volumetric analysis has shown great potential in addressing the shortcomings of linear tumor assessment, several challenges still impede clinical implementation, including limited data availability, variability in annotation, sensitivity to scanners and acquisition protocols, poor interpretability, and multimodal integration.

AI-Driven Tumor Characterization and Histological Subtype Classification in Lung Cancer Using CT Imaging

Background/Objectives: Lung cancer is still one of the top cancer mortality causes around the world, and there is a need for an accurate and clinically reliable diagnostic tool. While Computed Tomography (CT) imaging is very useful for evaluation of pulmonary nodules and tumor morphology, its interpretation is complicated by inter-patient variability, imaging artifacts, low tissue contrast, and tumor heterogeneity. Although Computer-Aided Diagnosis (CAD) systems have enhanced the diagnostic process, handcrafted feature-based approaches often fail to capture complex tumor characteristics, and numerous deep learning systems lack clinical interpretability. To tackle these challenges, this study suggests a unified diagnostic approach to characterize the tumor comprehensively. Methods: Lung window intensity clipping and the MedSAM foundation model are used to segment the tumor regions. After segmentation, handcrafted texture, shape, morphology and keypoint features are extracted in addition to deep features extracted by ResNet50. Particle Swarm Optimization (PSO) is used to select and refine the features, followed by an LSTM network that learns the sequential relationships among features for histological subtype classification. Results: It was observed that the proposed approach outperformed the benchmark approaches by attaining a higher accuracy of 93.70% and 94.70% on the Lung-PET-CT-Dx and LIDC-IDRI datasets, respectively. The ablation analysis supports the contribution of each module, clearly showing the progressive improvement of the overall classification performance obtained by integrating the complementary modules. Conclusions: The proposed framework effectively incorporated MedSAM-based tumor segmentation, radiomic feature analysis, and deep feature representation and sequential dependency modeling all in a single diagnostic workflow for lung cancer evaluation and diagnosis. These results prove its feasibility for explainable computer-aided diagnosis and decision support for lung cancer evaluation.