An Assessment on Implementation of Imperialist Competitive Algorithm for Motion Dataset Optimization at Radiotherapy with External Surrogates

One of the most important components in radiotherapy with external surrogates is utilizing consistent correlation model to estimate tumor location as model output on the basis of external markers motion dataset. In this study, imperialist competitive algorithm (ICA) was proposed to process and optimize motion dataset for correlation model. The simplicity of correlation model based on this algorithm may result in less targeting error with the least computational time.
A correlation model based on adaptive neuro-fuzzy inference system (ANFIS) was utilized with database of 20 patients treated with CyberKnife Synchrony system. In order to assess the effect of proposed data optimization algorithm, two strategies were considered. The correlation model was used with and without implementing ICA. Then, targeting error of ANFIS model was compared at two strategies using statistical analysis.
The results showed that implementing the proposed algorithm on ANFIS model could remarkably improve the performance accuracy of ANFIS correlation model by eliminating unnecessary and noisy inputs and making the model simpler. Moreover, model simplicity factor could highly reduce model computational time, which is attractive for clinical practice.
ICA was proposed as data optimization algorithm on motion dataset of patients treated with external surrogates’ radiotherapy. Our proposed algorithm could highly optimize the input motion dataset of correlation model for estimating tumor position by selecting enough data points with high degree of importance. The final results showed an improvement of targeting accuracy of correlation model, as well as a significant reduction at model computational time.