Machine learning approach using 18F-FDG-PET-radiomic features and the visibility of right ventricle 18F-FDG uptake for predicting clinical events in patients with cardiac sarcoidosis

<jats:title>Abstract</jats:title><jats:sec> <jats:title>Objectives</jats:title> <jats:p>To investigate the usefulness of machine learning (ML) models using pretreatment <jats:sup>18</jats:sup>F-FDG-PET-based radiomic features for predicting adverse clinical events (ACEs) in patients with cardiac sarcoidosis (CS).</jats:p> </jats:sec><jats:sec> <jats:title>Materials and methods</jats:title> <jats:p>This retrospective study included 47 patients with CS who underwent <jats:sup>18</jats:sup>F-FDG-PET/CT scan before treatment. The lesions were assigned to the training (<jats:italic>n</jats:italic> = 38) and testing (<jats:italic>n</jats:italic> = 9) cohorts. In total, 49 <jats:sup>18</jats:sup>F-FDG-PET-based radiomic features and the visibility of right ventricle <jats:sup>18</jats:sup>F-FDG uptake were used to predict ACEs using seven different ML algorithms (namely, decision tree, random forest [RF], neural network, k-nearest neighbors, Naïve Bayes, logistic regression, and support vector machine [SVM]) with tenfold cross-validation and the synthetic minority over-sampling technique. The ML models were constructed using the top four features ranked by the decrease in Gini impurity. The AUCs and accuracies were used to compare predictive performances.</jats:p> </jats:sec><jats:sec> <jats:title>Results</jats:title> <jats:p>Patients who developed ACEs presented with a significantly higher surface area and gray level run length matrix run length non-uniformity (GLRLM_RLNU), and lower neighborhood gray-tone difference matrix_coarseness and sphericity than those without ACEs (each, <jats:italic>p</jats:italic> < 0.05). In the training cohort, all seven ML algorithms had a good classification performance with AUC values of > 0.80 (range: 0.841–0.944). In the testing cohort, the RF algorithm had the highest AUC and accuracy (88.9% [8/9]) with a similar classification performance between training and testing cohorts (AUC: 0.945 vs 0.889). GLRLM_RLNU was the most important feature of the modeling process of this RF algorithm.</jats:p> </jats:sec><jats:sec> <jats:title>Conclusion</jats:title> <jats:p>ML analyses using <jats:sup>18</jats:sup>F-FDG-PET-based radiomic features may be useful for predicting ACEs in patients with CS.</jats:p> </jats:sec>