The evidence says that liver disease detection using CAD is one of the most efficient techniques but the presence of better organization of studies and the performance parameters to represent the result analysis of the proposed techniques are pointedly missing in most of the recent studies. Few benchmarked studies have been found in some of the papers as benchmarking makes a reader understand that under which circumstances their experimental results or outcomes are better and useful for the future implementation and adoption of the work. Liver diseases and image processing algorithms, especially in medicine, are the most important and important topics of the day. Unfortunately, the necessary data and data, as they are invoked in the articles, are low in this area and require the revision and implementation of policies in order to gather and do more research in this field. Detection with ultrasound is quite normal in liver diseases and depends on the physician's experience and skills. CAD systems are very important for doctors to understand medical images and improve the accuracy of diagnosing various diseases. In the following, we describe the techniques used in the various stages of a CAD system, namely: extracting features, selecting features, and classifying them. Although there are many techniques that are used to classify medical images, it is still a challenging issue for creating a universally accepted approach.