Enhancing Accuracy for the Detection of Liver Cancer Using Decision Tree in Comparison with Linear Contrast Filter

The proposed aim of the work is to improve the accuracy of automatic detection and classification of liver cancer. Materials and Methods: The sample size for each group in a study comparing the performance of Contrast Filter classifier with decision tree classifier was determined using GPower to be 10. The desired power for the study is 0.80 and the alpha level is 0.08. For performance analysis, 70% of the images will be used for training and 30% of the images will be used for testing and validation. Result: The Contrast Filter algorithm achieved an accuracy of 72.32% which was higher than the accuracy of 78.71% achieved by the decision tree algorithm. An independent sample T-test was conducted to compare the results and found that the difference in accuracy was statistically significant (2 tailed) (p = 0.004). Conclusion: The Contrast Filter algorithm outperforms with higher accuracy compared to the decision tree algorithm.