Dr. Hang See Pheng
Department of Mathematical Sciences
Faculty of Science
Universiti Teknologi Malaysia

Massive Training Artificial Immune Recognition System for Lung Nodules Detection

In the early detection and diagnosis of lung nodule, computer aided detection (CAD) has became crucial to assist the radiologists in interpreting medical images and decision making. However, some limitations have been found in the existing CAD algorithms for detecting lung nodules such as imprecision classification due to the inaccurate of segmentation and expensive computation time.  The Massive Training Artificial Immune Recognition System (MTAIRS) is developed to detect the lung nodules on Computed Tomography (CT) scans based on the pixel machine learning and artificial immune based system-Artificial Immune Recognition System (AIRS). There are some mathematical knowledge and equations have been applied in the research. The segmentation and feature calculation are not needed in the pixel-based machine learning to avoid the information lost during the image processing. The MTAIRS is found reducing the computation time and accomplishing good accuracy in the detection of lung nodules on CT scans compare to the other well-known pixelbased classification algorithms. To further provide comparative analysis of pixelbased classification algorithms in lung nodules detection, a pixel-based evaluation method of Kullback Leibler (KL) divergence is developed in this study.