This article was initially printed right here
Annu Int Conf IEEE Eng Med Biol Soc. 2021 Nov;2021:3374-3377. doi: 10.1109/EMBC46164.2021.9630988.
ABSTRACT
In the final many years, MRI was confirmed a great tool for the analysis and characterization of Prostate Cancer (PCa). In the literature, many research targeted on characterizing PCa aggressiveness, however a number of have distinguished between low-aggressive (Gleason Grade Group (GG) <=2) and high-aggressive (GG>=3) PCas primarily based on biparametric MRI (bpMRI). In this research, 108 PCas have been collected from two completely different facilities and have been divided into coaching, testing, and validation set. From Apparent Diffusion Coefficient (ADC) map and T2-Weighted Image (T2WI), we extracted texture options, each 3D and 2D, and we carried out three completely different strategies of Feature Selection (FS): Minimum Redundance Maximum Relevance (MRMR), Affinity Propagation (AP), and Genetic Algorithm (GA). From the ensuing subsets of predictors, we educated Support Vector Machine (SVM), Decision Tree, and Ensemble Learning classifiers on the coaching set, and we evaluated their prediction capacity on the testing set. Then, for every FS methodology, we selected one of the best classifier, primarily based on each coaching and testing performances, and we additional assessed their generalization functionality on the validation set. Between the three greatest fashions, a Decision Tree was educated utilizing solely two options extracted from the ADC map and chosen by MRMR, attaining, on the validation set, an Area Under the ROC (AUC) equal to 81%, with sensitivity and specificity of 77% and 93%, respectively.Clinical Relevance- Our greatest mannequin demonstrated to have the ability to distinguish low-aggressive from high-aggressive PCas with high accuracy. Potentially, this method might assist clinician to noninvasively distinguish between PCas which may want energetic therapy and those who might probably profit from energetic surveillance, avoiding biopsy-related problems.
PMID:34891963 | DOI:10.1109/EMBC46164.2021.9630988
