Lung diseases present a significant challenge in medical diagnostics due to their diverse etiologies and overlapping clinical symptoms, necessitating advanced classification methods to enhance diagnostic accuracy and treatment efficacy. In recent years, Machine Learning algorithms have increasingly been employed to enhance the diagnosis of medical images, offering more accurate, efficient, and reliable results. This study focuses on multi-class lung disease classification using CT scans, leveraging advanced texture descriptors including Gabor filters, Gray-Level Co-occurrence Matrix (GLCM), Local Binary Patterns (LBP), and Histogram of Oriented Gradients (HOG) for feature extraction. These features are integrated with the DenseNet121 architecture, which is also employed for extracting deep features from CT images. Various ML algorithms, such as Random Forest (RF), Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), Decision Tree (DT), Naive Bayes (NB), and K-Nearest Neighbors (KNN) are employed for classification. Number of experiments have been conducted to assess the performance of these models both with individual texture descriptors and their combinations. Results demonstrate that incorporating texture descriptors alongside deep features significantly improves predictive accuracy. The proposed model achieved accuracy of 99.92%, precision, recall and F1-score of 99.94%. The study shows that combining different texture descriptors further enhances the metrics for detecting and diagnosing multiple lung diseases, leading to more precise and robust classification outcomes.