Neutrosophic Topological Spaces for Lung Cancer  Detection in Chest X-Rays: A Novel Approach to  Uncertainty Management

A. A. Salama; Doaa E. Mossa; Mahmoud Y. Shams; Huda E. Khalid; Ahmed K. Essa

Authors

A. A. Salama Department of Mathematics and Computer Science, Faculty of Sciences, Port Said University, Port Said, 42526, Egypt.
Doaa E. Mossa High Institute of Computer Science and Information Systems, Delta Acadmey,City, Dakahlia, Egypt
Mahmoud Y. Shams Faculty of Artificial Intelligence, Kafrelsheikh University, Kafrelsheikh, 33511, Egypt
Huda E. Khalid Telafer University, The Administration Assistant for the President of the Telafer University, Telafer, Iraq;
Ahmed K. Essa Telafer University, The Administration Assistant for the President of the Telafer University, Telafer, Iraq

Keywords:

: Neutrosophic Sets, Neutrosophic Topological Spaces, Medical Diagnosis, Lung Cancer Detection, Chest X-Ray Images, Uncertainty, Decision-Making

Abstract

Decision-making in medical diagnosis is often hampered by uncertainties due to
incomplete, ambiguous, and evolving information. In reviewing the traditional methods for lung
cancer detection, we found that crisp and logic values have more difficulties and challenges. These
challenges related to the big data analytics, uncertainty values, and the different circumstances that
make it harder for prediction. In this work, we propose a novel approach that use a Neutrosophic
Topological Spaces (NTS) for the lung cancer detection in the chest X-ray images. Furthermore, the
proposed NTS leverage the strengths points of Neutrosophic Sets (NS) which include the degrees of
truth (T), indeterminacy (I), and falsity (F). The proposed model provides more informative results
about the uncertainty cases compared with the traditional methods. The results indicated that the
proposed NTS approach achieved highest accuracy reached to 85.5% with a sensitivity 88.2%,
specificity 82.1%, and AUC 0.91. which mean that the proposed NTS approach are more reliable and
efficient than traditional methods for uncertainty.

DOI: 10.5281/zenodo.14172142