Radiomics and artificial intelligence have the potential to become a valuable tool in clinical applications. Frequently, radiomic analyses through machine learning methods present issues caused by high dimensionality and multicollinearity, and redundant radiomic features are usually removed based on correlation analysis. We assessed the effect of preprocessing—in terms of voxel size resampling, discretization, and filtering—on correlation-based dimensionality reduction in radiomic features from cardiac T1 and T2 maps of patients with hypertrophic cardiomyopathy. For different combinations of preprocessing parameters, we performed a dimensionality reduction of radiomic features based on either Pearson’s or Spearman’s correlation coefficient, followed by the computation of the stability index. With varying resampling voxel size and discretization bin width, for both T1 and T2 maps, Pearson’s and Spearman’s dimensionality reduction produced a slightly different percentage of remaining radiomic features, with a relatively high stability index. For different filters, the remaining features’ stability was instead relatively low. Overall, the percentage of eliminated radiomic features through correlation-based dimensionality reduction was more dependent on resampling voxel size and discretization bin width for textural features than for shape or first-order features. Notably, correlation-based dimensionality reduction was less sensitive to preprocessing when considering radiomic features from T2 compared with T1 maps.

Substantial reduction in the thyroid volume (up to 70-80%) after ¹³¹I therapy of Graves' disease has been demonstrated and reported in the literature. Recently a mathematical model of thyroid mass reduction during the first month after therapy has been developed and a new algorithm for the radiation committed dose calculation has been proposed. Reduction of the thyroid mass and the radiation committed dose to the gland depend on a parameter k, defined for each subject. The calculation of k allows the prediction of the activity to administer, depending on the radiation committed dose chosen by the physician. In this paper a method for calculating k is proposed. The calculated values of k are compared to values derived from measurements of the changes in thyroid mass in twenty-six patients treated by ¹³¹I for Graves' disease. The radiation committed dose to the thyroid can be predicted within 21%, and the radioiodine activity to administer to the patient can be predicted within 22% using the calculated values of k. The thyroid volume reduction during the first month after therapy administration can be also predicted with good accuracy using the calculated values of k. The radiation committed dose and the radioiodine activity to administer were calculated using a new, very simple algorithm. A comparison between the values calculated by this new algorithm and the old, classical Marinelli-Quimby algorithm shows that the new method is more accurate.