CCSA-RF enhanced lightweight multi-scale CNN framework for robust lung cancer classification
Year: 2026
Type: article
Abstract: The early and accurate diagnosis of lung cancer is an urgent issue owing to the noise of imaging, nonhomogeneous nodule structures, and high cost of computing the current deep learning architectures. Most modern methods find it difficult to strike a balance between diagnostic accuracy, strength, and efficiency, especially with multimodal medical images. To overcome these shortcomings, this study proposes a single lightweight-based learning framework that combines preprocessing using wavelets, feature extraction under attention, chaos-based feature selection, and multiscale convolutional learning. First, a Ricker wavelet invariant center-weighted mean (RWICWM) filter was used to reduce noise without unimportant structures in the diagnostics. An Attention-Based DenseNet (ATT-DenseNet) is then used to extract discriminative features by highlighting the parts of the image that contain cancer-related features with channel-wise attention. The chaotic convolutional spectral analysis random forest (CCSA-RF) mechanism was proposed to select informative features and eliminate redundancy. Finally, a lightweight multi-scale CNN (LMS-CNN) is employed to achieve effective and precise classification of lung cancer. On the LC25000 and LIDC-IDRI datasets, the proposed framework attained an accuracy of 96.54, recall of 96.79, precision of 96.90, and F1-score of 96.32, with both datasets reporting similar results. The findings indicate that the proposed method enhances the classification accuracy and strength and still has a low computational complexity, which renders it applicable for real-world clinical use. The proposed research offers a computationally and theoretically sound solution that sets the state-of-the-art in lung cancer diagnostics. • This study proposes an advanced algorithm that combines two deep neural networks to facilitate early lung cancer diagnosis. The classification method is based on a lightweight multiscale convolutional neural network (LMS-CNN) and uses a Ricker wavelet-invariant central weighted mean (RWICWM) filter during preprocessing to reduce noise, improve nodule classification, and preserve important boundaries and textures in medical images. • The Attention-Based Dense Network (ATT-DenseNet) improves the identification of the most discriminative regions of lung nodules and enhances representation learning. Additionally, ATT-DenseNet allows the model to concentrate on specific image areas by assigning higher weighted values. The subsequent stage employs a chaotic convolutional search and random forest (CCSARF) method to represent features such as diameter, edge, spike, lobe, microcalcification, and malignant tumors. • Applying these methods to the LC25000 dataset enabled us to achieve 96.54%, 96.79%, 96.32% F1-score, and 96.90%, respectively. The model demonstrated good performance in identifying the most common cancer types across five classifications and two subtypes of squamous cell carcinoma. The model's convolutional neural network (LW-MS) architecture contains 1.1 million learnable parameters and can be applied in real time, reducing processing time and improving system efficiency.
Source: Intelligence-Based Medicine
Authors: R. Gayathri, R. Thilagavathy
Institution: SRM Institute of Science and Technology
Language: English
Topic: AI in cancer detection
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,420
OpenAlex ID: W7130863003