Automated multiparametric localization of prostate cancer based on B-mode, shear-wave elastography, and contrast-enhanced ultrasound radiomics

Abstract

Objectives: The aim of this study was to assess the potential of machine learning based on B-mode, shear-wave elastography (SWE), and dynamic contrast-enhanced ultrasound (DCE-US) radiomics for the localization of prostate cancer (PCa) lesions using transrectal ultrasound.

Methods: This study was approved by the institutional review board and comprised 50 men with biopsy-confirmed PCa that were referred for radical prostatectomy. Prior to surgery, patients received transrectal ultrasound (TRUS), SWE, and DCE-US for three imaging planes. The images were automatically segmented and registered. First, model-based features related to contrast perfusion and dispersion were extracted from the DCE-US videos. Subsequently, radiomics were retrieved from all modalities. Machine learning was applied through a random forest classification algorithm, using the co-registered histopathology from the radical prostatectomy specimens as a reference to draw benign and malignant regions of interest. To avoid overfitting, the performance of the multiparametric classifier was assessed through leave-one-patient-out cross-validation.

Results: The multiparametric classifier reached a region-wise area under the receiver operating characteristics curve (ROC-AUC) of 0.75 and 0.90 for PCa and Gleason > 3 + 4 significant PCa, respectively, thereby outperforming the best-performing single parameter (i.e., contrast velocity) yielding ROC-AUCs of 0.69 and 0.76, respectively. Machine learning revealed that combinations between perfusion-, dispersion-, and elasticity-related features were favored.

Conclusions: In this paper, technical feasibility of multiparametric machine learning to improve upon single US modalities for the localization of PCa has been demonstrated. Extended datasets for training and testing may establish the clinical value of automatic multiparametric US classification in the early diagnosis of PCa.

Key points: • Combination of B-mode ultrasound, shear-wave elastography, and contrast ultrasound radiomics through machine learning is technically feasible. • Multiparametric ultrasound demonstrated a higher prostate cancer localization ability than single ultrasound modalities. • Computer-aided multiparametric ultrasound could help clinicians in biopsy targeting.

Keywords: Contrast media; Elasticity imaging techniques; Machine learning; Prostate cancer; Ultrasonography.

Fig. 1

Schematic overview of the proposed classification framework, with information from shear-wave elastography and contrast-enhanced ultrasound recording shown in blue and red, respectively

Fig. 2

Correlation matrix of the derived radiomics in terms of the linear Pearson correlation coefficient; correlations that are not significantly (p > 0.05) reflected by a linear correlation are indicated by a black square

Fig. 3

Image plane example, showing the B-mode (a), Young’s modulus (SWE) (b), Péclet number (c), spatiotemporal correlation (d), dispersion-related parameter (e), wash-in time (f), velocity (g), velocity relative to image median (h), 2-mm entropy of velocity (i), and resulting multiparametric map (j). In each map, the prostate and zonal segmentations are depicted in white, the calcifications are encircled in blue, and histopathologically confirmed malignant and benign ROIs are indicated in red and green, respectively

Fig. 4

Overview of the frequency at which radiomics are selected for the highest-order branches among all trees in the forest. Radiomics are grouped according to the model-based parameters

Fig. 5

Overview of the parameter values and classifier score for the velocity (a), Young’s modulus (E) (b), and the multiparametric classifier score (c). Individual regions are represented by a bullet. The violin plots represent the group distribution in the PZ (left, blue) and TZ (right, red). Significant and highly significant differences according to a Wilcoxon rank sum test are indicated with a single asterisk and double asterisks