Material Decomposition Using Ensemble Learning for Energy-Resolved Computed Tomography

Material decomposition facilitates the differentiation of different materials in X-ray imaging. As an alternative to the previous empirical material decomposition methods, we performed material decomposition using ensemble learning methods in this work. Three representative ensemble methods with two decision trees as the base learning algorithms were implemented to perform material decomposition in both simulation study and experimental study. The results were quantitatively evaluated for comparison study. The performance of the base learning algorithms was improved by using appropriate ensemble methods. The results indicate that it is feasible and promising to perform material decomposition using ensemble learning, which is valuable to be further investigated.