Texture analysis of CT-images for early detection of liver malignancy

In certain medical images e.g., ultrasound it has been seen that the texture conveys useful diagnostic information but in CT and MR images its proper use has not been established. Present study is an attempt to investigate the use of texture analysis for early detection of liver malignancy when the onset of disease is beyond human perception, using CT-images. Using grey level run length as a primitive, five parameters viz., Short Run Emphasis (SRE), Long Run Emphasis (LRE), Grey Level Distribution (GLD), Run Length Distribution (RLD) and Run Percentage (RP) have been studied for this purpose. It has been found that the GLD feature, obtained using Grey Level Run Length Method (GLRLM) conveys useful information about the onset of this disease with a confidence level of above 99 percent. The results have been confirmed on the basis of clinical studies.