Liver Detection Based on Iridology using Local Binary Pattern Extraction

Iridology is one of the technologies that is used in medical sector for helping the medical staffs in analyzing the patient’s health from the observed iris. The assessment is based on mapping that uses iridology chart or Jensen diagram. In iris-mapping-process, the location representing each organ is close to each other, thus it is quite difficult to distinguish the normal and abnormal characteristics of the organs. This factor has the potential to cause some errors in analyzing the iris. Thus a computerized system with image processing techniques will help to detect damaged fiber networks more accurately. Therefore, this study will be focused on diagnosing the liver condition based on the iris-mapping image of the right eye, by using the eXtended Center Symmetric Local Binary Pattern (XCS-LBP) method. The XCS-LBP method is used due to its extraction method feature, which is known to be more robust towards the changes of light and noises. Hence, the XCSLBP method is able to extract the features more clearly in order to get normal and abnormal iris characteristics. The result shows that the XCS-LBP method provides 92% of accuracy, 83% of sensitivity, and 100% of specificity and precision other than that of the compared methods.