Cancer is generally defined as the uncontrollable increase of number of cells in the body. These cells might be formed anywhere in the body and spread to other parts of the body. Although the mortality rate of cancer is high, it is possible to decrease cancer cases by up to 30% to 50% through taking a healthy lifestyle and avoiding unhealthy habits. Imaging is one of the powerful technologies used for detecting and treating cancer at its early stages. Nowadays, scientists admit that medical images hold more information than their diagnosis, which is called a radiomics approach. Radiomics demonstrate that images comprise numerous quantitative features that are useful in predicting, detecting, and treating cancers in a personalized manner. While radiomics can extract numerous features, not all of them are useful. It should not be neglected that the outcome of data analysis is highly dependent on the selected features. There are different ways of finding the most reliable features. One possible way is to select all extracted features, analyze them, and find the most reproducible and reliable ones. Different statistical analysis metrics could analyze the features. To discover and introduce the most accurate metrics, in this paper, different statistical metrics used for measuring the stability and reproducibility of the features are investigated.