Performance enhancement for disease diagnostic systems has been utmost challenging aspect of providing further treatments or proceeds surgeries without any possible delays. In recent times, various data mining techniques are being applied as the ratio of lung cancer is increasing enormously in recent years and require significant developments in its accurate detection at a possible early stage to cure the patients from further suffering. Developing a diagnostic system for lung cancer demands efficiency in processing and classification of X-rays of normal and cancerous cases. In this work, robust computer aided diagnostic system is proposed by utilizing modified clustering based classifiers such as Support Vector Machine (SVM) and k-Nearest Neighbors (k-NN) with optimized processing techniques for feature processing and selection of suitable features to enhance system's performance in terms of accuracy, sensitivity and specificity. Overall, this work has proved to have a maximum detection rate with respect to earlier techniques applied. In future this approach will be implemented for determining the region of interest (ROI) and classifying the severity of cancer cases as mild, moderate or critical.