Statistical Analysis of Law’s Mask Texture Features for Cancer and Water Lung Detection

Lung cancer is distinguished by presenting one of the highest rates of mortality. Detecting and curing the disease in the early stages provides patients with a high chance of survival. Moreover, the presence of an excessive amount of water in lung is usually accompanied by a high mortality rate. Thus, there is an urge to develop an automatic technique for detecting and monitoring lung water. This paper reports a study conducted on the use of Laws’ masks to calculate energy statistics that gives description features of a cancerous and water lung texture that can be used in turn for texture discrimination. Laws’ masks method has been recognized as a very useful tool in image processing for texture analysis, however it has not been utilized in cancer or water lung characterization. The proposed algorithm proceeds in three steps: image preprocessing, lung region extraction and texture feature extraction. To reduce the feature space, statistic t-test and its p values for feature selection are proposed. DICOM CT images are used to test the proposed algorithm. Experimental results show that Laws’ method has high capability to extract texture features that can discriminate between cancer and normal cases, water and normal cases as well as cancer and water cases.