Lung Cancer Detection in Histopathological Images Using Lion Optimization Algorithm and Machine Learning Classifiers

Lung cancer, a common malignancy in both men and women, requires early detection to reduce its severe impact. However, symptoms usually emerge in advanced stages, necessitating medical imaging for timely detection. Scientists have designed algorithms to identify lung cancer in histopathological scans. This study utilizes standard histopathological images from the Kaggle database. A computer-aided detection approach is implemented, beginning with an adaptive median filter for noise reduction and Contrast Limited Adaptive Histogram Equalization (CLAHE) to improve contrast in histopathological images. The Lion Optimization Algorithm (LOA) is then applied for feature extraction, where the extracted image intensity values serve as inputs for classification. Five classifiers—Gaussian Mixture Model (GMM), Naïve Bayes Classifier (NBC), Logistic Regression (LR), Support Vector Machine (SVM), Multilayer Perceptron (MLP), and K-Nearest Neighbor (KNN)—are employed to distinguish between Benign and Squamous Cell Carcinoma (SCC) cases. The findings reveal that the MLP classifier, integrated with LOA, attains a superior accuracy of 97.135%, as measured by established benchmark criteria.