Back to all papers

An innovative deep learning paradigm for automated detection and accurate classification of lung nodules in magnetic resonance imaging.

June 22, 2026pubmed logopapers

Authors

Ul Haq E,Ting FF,Phan RC,Ting CM

Affiliations (3)

  • SDAIA-KFUPM Joint Research Center for Artificial Intelligence, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia.
  • School of Information Technology, Monash University, Malaysia Campus, Malaysia.
  • School of Information Technology, Monash University, Malaysia Campus, Malaysia. Electronic address: [email protected].

Abstract

Lung cancer, the predominant kind of cancer, needs considerable care, since inadequate treatment may lead to fatal outcomes. The integration of computer-aided diagnostic (CAD) systems is crucial for the early detection of lung nodules, greatly aiding in the decrease of death rates associated with lung cancer. Magnetic resonance imaging (MRI) is a proficient technique for diagnosing lung cancer. Diverse techniques have been investigated for the detection of lung nodules in computed tomography (CT) images. The diagnosis process is greatly influenced by the physician's expertise, which may lead to the unintentional neglect of specific patients and ensuing consequences. Deep learning has become a significant and well-established method in several fields of diagnostic medical imaging. This study introduces a deep learning approach for identifying lung nodules in magnetic resonance imaging. Automatic feature extraction and classification are accomplished via the use of a deep convolutional neural network (CNN) of 11-layers. A strategy for intensity normalization is established as a preprocessing measure, which, when integrated with data augmentation techniques, has significant effectiveness in identifying and categorizing benign and malignant lung nodules. The assessment of the suggested methodology included 243T2-weighted MR images acquired from the First Affiliated Hospital of Shenzhen University. The experimental findings provide a diagnostic accuracy of 98.5% and a Dice similarity coefficient (DSC) of 97.1% in differentiating benign from malignant lung nodules. The findings indicate that the newly designed model exceeds existing state-of-the-art approaches in terms of accuracy and operational efficiency.

Topics

Journal Article

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.