An Optimized Clinical Framework for Lung Cancer Classification from CT Images

Read full text Summarise with VIRAT
Authors: Eslam Mahmoud Fouda, Ahmed Mohamed Saad Emam Saad, Ahmed Elrefaey
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Lung cancer remains the leading cause of cancer deaths worldwide, thereby necessitating the development of accurate, efficient, and automated diagnostic tools to identify the disease at an early stage. This work presents a novel hybrid deep learning framework for multi-class lung cancer classification from chest CT scans. This framework effectively combines Gabor-based texture optimization, MobileNetV1 for transfer learning-based feature extraction, and an optimized Support Vector Machine (SVM) to ensure robust classification accuracy. Four types of clinically reasonable data augmentation were carefully performed, and the optimized SVM classifier was rigorously evaluated on a test set not used during training. This classifier achieved an accuracy of 99.68%, significantly better than the baseline SVM configuration (99.36%). The model demonstrates superior potential in identifying aggressive subtypes (e.g., 100% recall for large-cell carcinoma) as well as substantially reducing false positive rates. Furthermore, thanks to its sub-second inference time and lightweight architecture, this system can be easily deployed and implemented in resource-constrained clinical settings.

Fields of Study & Subjects

AI in cancer detection Radiomics and Machine Learning in Medical Imaging Lung Cancer Diagnosis and Treatment

Publication details

Journal / source
International Journal of Telecommunications
Volume
06
Issue
01
Pages
1–13
Publication date
27 March 2026
Language
EN