Prediction of recurrence after catheter ablation for atrial fibrillation using left atrial morphology on preprocedural computed tomography: application of radiomics

Background Radiomics is a comprehensive analysis methodology of medical image and involves the extraction of numerous features from standard imaging. Its usefulness has been reported mainly in the field of cancer for diagnosis and prediction of prognosis. In the territory of cardiac imaging, several reports have investigated the utility of radiomics for classifying the risk of prognosis in coronary artery disease, and few practical applications have been reported for patients with atrial fibrillation (AF) who underwent pulmonary vein isolation (PVI). Although the left atrial morphology can affect the clinical course after the PVI procedure, it is unclear whether the radiomics feature values of the left atrial morphology on cardiac computed tomography (CT) is useful for predicting the AF recurrence after PVI. Purpose To predict the recurrence of AF after PVI using the radiomics feature values of the left atrial morphology on cardiac computed tomography (CT). Methods We analyzed 525 consecutive three-dimensional cardiac CT in patients with atrial fibrillation who underwent PVI from 2018 to 2019 in our institute. After marking the region of interest on left atrium (including the root of pulmonary veins) semiautomatically, 107 radiomics feature values were obtained by Python program. After excluding the parameters having collinearity or with low predictive capability for the recurrence of AF after PVI, 42 parameters were applied to the final prediction model. Two prediction models were constructed by multivariate Cox regression analysis and machine learning model by support vector machine algorithm. Results The area under the curve (AUC) for predicting the recurrence of AF was 0.815 for the multivariate Cox regression model and 0.826 for the machine learning model by support vector machine. Conclusion The radiomics feature values on preprocedural cardiac CT could be helpful for predicting the recurrence of AF after PVI. Since radiomics feature analysis yields a huge number of numerical values representing the left atrial morphology in a reproducible manner, it would provide a new direction to construct a good prediction model using machine learning including artificial intelligence out of a routine cardiac CT scan. Funding Acknowledgement Type of funding sources: None.