Multiparametric Ultrasound Imaging of Prostate Cancer Using Deep Neural Networks.

OBJECTIVE: A deep neural network (DNN) was trained to generate a multiparametric ultrasound (mpUS) volume from four input ultrasound-based modalities (acoustic radiation force impulse [ARFI] imaging, shear wave elasticity imaging [SWEI], quantitative ultrasound-midband fit [QUS-MF], and B-mode) for the detection of prostate cancer. METHODS: A DNN was trained using co-registered ARFI, SWEI, MF, and B-mode data obtained in men with biopsy-confirmed prostate cancer prior to radical prostatectomy (15 subjects, comprising 980,620 voxels). Data were obtained using a commercial scanner that was modified to allow user control of the acoustic beam sequences and provide access to the raw image data. For each subject, the index lesion and a non-cancerous region were manually segmented using visual confirmation based on whole-mount histopathology data. RESULTS: In a prostate phantom, the DNN increased lesion contrast-to-noise ratio (CNR) compared to a previous approach that used a linear support vector machine (SVM). In the in vivo test datasets (n = 15), the DNN-based mpUS volumes clearly portrayed histopathology-confirmed prostate cancer and significantly improved CNR compared to the linear SVM (2.79 ± 0.88 vs. 1.98 ± 0.73, paired-sample t-test p < 0.001). In a sub-analysis in which the input modalities to the DNN were selectively omitted, the CNR decreased with fewer inputs; both stiffness- and echogenicity-based modalities were important contributors to the multiparametric model. CONCLUSION: The findings from this study indicate that a DNN can be optimized to generate mpUS prostate volumes with high CNR from ARFI, SWEI, MF, and B-mode and that this approach outperforms a linear SVM approach.

Prostate Cancer Detection Using 3-D Shear Wave Elasticity Imaging.

Transrectal ultrasound (TRUS) B-mode imaging provides insufficient sensitivity and specificity for prostate cancer (PCa) targeting when used for biopsy guidance. Shear wave elasticity imaging (SWEI) is an elasticity imaging technique that has been commercially implemented and is sensitive and specific for PCa. We have developed a SWEI system capable of 3-D data acquisition using a dense acoustic radiation force (ARF) push approach that leads to enhanced shear wave signal-to-noise ratio compared with that of the commercially available SWEI systems and facilitates screening of the entire gland before biopsy. Additionally, we imaged and assessed 36 patients undergoing radical prostatectomy using 3-D SWEI and determined a shear wave speed threshold separating PCa from healthy prostate tissue with sensitivities and specificities akin to those for multiparametric magnetic resonance imaging fusion biopsy. The approach measured the mean shear wave speed in each prostate region to be 4.8 m/s (Young's modulus E = 69.1 kPa) in the peripheral zone, 5.3 m/s (E = 84.3 kPa) in the central gland and 6.0 m/s (E = 108.0 kPa) for PCa with statistically significant (p < 0.0001) differences among all regions. Three-dimensional SWEI receiver operating characteristic analyses identified a threshold of 5.6 m/s (E = 94.1 kPa) to separate PCa from healthy tissue with a sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and area under the curve (AUC) of 81%, 82%, 69%, 89% and 0.84, respectively. Additionally, a shear wave speed ratio was assessed to normalize for tissue compression and patient variability, which yielded a threshold of 1.11 to separate PCa from healthy prostate tissue and was accompanied by a substantial increase in specificity, PPV and AUC, where the sensitivity, specificity, PPV, NPV and AUC were 75%, 90%, 79%, 88% and 0.90, respectively. This work illustrates the feasibility of using 3-D SWEI data to detect and localize PCa and demonstrates the benefits of normalizing for applied compression during data acquisition for use in biopsy targeting studies.

Deep Convolutional Neural Networks for Displacement Estimation in ARFI Imaging.

Ultrasound elasticity imaging in soft tissue with acoustic radiation force requires the estimation of displacements, typically on the order of several microns, from serially acquired raw data A-lines. In this work, we implement a fully convolutional neural network (CNN) for ultrasound displacement estimation. We present a novel method for generating ultrasound training data, in which synthetic 3-D displacement volumes with a combination of randomly seeded ellipsoids are created and used to displace scatterers, from which simulated ultrasonic imaging is performed using Field II. Network performance was tested on these virtual displacement volumes, as well as an experimental ARFI phantom data set and a human in vivo prostate ARFI data set. In the simulated data, the proposed neural network performed comparably to Loupas's algorithm, a conventional phase-based displacement estimation algorithm; the rms error was [Formula: see text] for the CNN and 0.73 [Formula: see text] for Loupas. Similarly, in the phantom data, the contrast-to-noise ratio (CNR) of a stiff inclusion was 2.27 for the CNN-estimated image and 2.21 for the Loupas-estimated image. Applying the trained network to in vivo data enabled the visualization of prostate cancer and prostate anatomy. The proposed training method provided 26 000 training cases, which allowed robust network training. The CNN had a computation time that was comparable to Loupas's algorithm; further refinements to the network architecture may provide an improvement in the computation time. We conclude that deep neural network-based displacement estimation from ultrasonic data is feasible, providing comparable performance with respect to both accuracy and speed compared to current standard time-delay estimation approaches.

Multiparametric Ultrasound for Targeting Prostate Cancer: Combining ARFI, SWEI, QUS and B-Mode.

Diagnosing prostate cancer through standard transrectal ultrasound (TRUS)-guided biopsy is challenging because of the sensitivity and specificity limitations of B-mode imaging. We used a linear support vector machine (SVM) to combine standard TRUS imaging data with acoustic radiation force impulse (ARFI) imaging data, shear wave elasticity imaging (SWEI) data and quantitative ultrasound (QUS) midband fit data to enhance lesion contrast into a synthesized multiparametric ultrasound volume. This SVM was trained and validated using a subset of 20 patients and tested on a second subset of 10 patients. Multiparametric US led to a statistically significant improvements in contrast, contrast-to-noise ratio (CNR) and generalized CNR (gCNR) when compared with standard TRUS B-mode and SWEI; in contrast and CNR when compared with MF; and in CNR when compared with ARFI. ARFI, MF and SWEI also outperformed TRUS B-mode in contrast, with MF outperforming B-mode in CNR and gCNR as well. ARFI, although only yielding statistically significant differences in contrast compared with TRUS B-mode, captured critical qualitative features for lesion identification. Multiparametric US enhanced lesion visibility metrics and is a promising technique for targeted TRUS-guided prostate biopsy in the future.