EFFIGBM-NET FOR AUTOMATED PULMONARY NODULE CLASSIFICATION

Abstract

Lung cancer is still one of the major causes of cancer death in the world, and early detection of pulmonary nodules is a key to improving the survival rate. This study introduces a machine learning-based framework to classify CT images into benign, malignant and normal categories called PNCA (Pulmonary Nodule Classifier Algorithm). Given the accessibility, affordability, and effectiveness of computed tomography (CT) imaging for pulmonary health assessment, it was chosen. The ANOVA F-ratio method was used to identify the discriminative attributes in the process of feature importance and selection. Later, Support Vector Machine (SVM), Random Forest, and Multi-Layer Perceptron classifiers were trained and optimized by using grid and randomized search methods for hyperparameters. The experimental results showed that the proposed framework achieved the highest classification accuracy of 99.09%, which indicated the stability and reliability of the proposed framework. The study shows that machine learning-enabled CT analysis has the potential to assist in clinical decision-making, early diagnosis, and help lower lung cancer mortality, especially in the context of resource-limited healthcare settings that reflect sustainable healthcare goals in the world.