Modied tumor diagnosis by classication and use of canonical correlation and support vector machines methods

The main objective of this research is to investigate techniques for classifying tumor grades based on image processing. The algorithms used to classify tumors are introduced, and their performance in the experimental results are evaluated. In the proposed algorithm, rst, the scan images of the lung are pre-processed and then, the histogram, texture, and geometric features are extracted. These characteristics are then employed in Support Vector Machines (SVM) and Canonical Correlation Analysis (CCA) classiers to diagnose tumors and classify benign and malignant types. These integrated approaches to investigation of medical images are vital tools for improving the diagonalization accuracy. In the current research, experimental and simulated medical images are employed. The outcomes of the developed techniques in this research are compared with those found in the literature review to conrm the ecacy and reliability of the proposed approach in diagnosing and classifying tumors. In addition to high accuracy in diagnosis, this method is also a low-cost and low-risk method. Owing to its very high sensitivity, this method has the desired values of two criteria of precision and specicity as well as the small number of features used for classication; therefore, the developed method was proposed as an ecient and appropriate one for tumor classication.