A model-based dynamic MRI called k-t BLAST/SENSE has drawn significant attention from the MR imaging community because of its improved spatio-temporal resolution. Recently, we showed that the k-t BLAST/SENSE corresponds to the special case of a new dynamic MRI algorithm called k-t FOCUSS that is optimal from a compressed sensing perspective. The main contribution of this article is an extension of k-t FOCUSS to a more general framework with prediction and residual encoding, where the prediction provides an initial estimate and the resid-

We demonstrate the possibility of a domain-adaptation when only a limited amount of MR data is available. The proposed method surpasses the existing compressed sensing algorithms in terms of the image quality and computation time.

Multi-parametric MRI (mp-MRI) is considered the best non-invasive imaging modality for diagnosing prostate cancer (PCa). However, mp-MRI for PCa diagnosis is currently limited by the qualitative or semi-quantitative interpretation criteria, leading to inter-reader variability and a suboptimal ability to assess lesion aggressiveness. Convolutional neural networks (CNNs) are a powerful method to automatically learn the discriminative features for various tasks, including cancer detection. We propose a novel multi-class CNN, FocalNet, to jointly detect PCa lesions and predict their aggressiveness using Gleason score (GS). FocalNet characterizes lesion aggressiveness and fully utilizes distinctive knowledge from mp-MRI. We collected a prostate mp-MRI dataset from 417 patients who underwent 3T mp-MRI exams prior to robotic-assisted laparoscopic prostatectomy (RALP). FocalNet is trained and evaluated in this large study cohort with 5-fold cross-validation. In the free-response receiver operating characteristics (FROC) analysis for lesion detection, FocalNet achieved 89.7% and 87.9% sensitivity for index lesions and clinically significant lesions at 1 false positive per patient, respectively. For GS classification, evaluated by the receiver operating characteristics (ROC) analysis, FocalNet received the area under the curve (AUC) of 0.81 and 0.79 for the classifications of clinically significant PCa (GS≥3+4) and PCa with GS≥4+3, respectively. With the comparison to the prospective performance of radiologists using the current diagnostic guideline, FocalNet demonstrated comparable detection sensitivity for index lesions and clinically significant lesions, only 3.4% and 1.5% lower than highly experienced radiologists without statistical significance.

No abstract

Purpose: To measure and characterize variations in the transmitted radio frequency (RF) (B1ϩ) field in cardiac magnetic resonance imaging (MRI) at 3 Tesla. Knowledge of the B1ϩ field is necessary for the calibration of pulse sequences, image-based quantitation, and signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) optimization. Materials and Methods:A variation of the saturated double-angle method for cardiac B1ϩ mapping is described. A total of eight healthy volunteers and two cardiac patients were scanned using six parallel short-axis slices spanning the left ventricle (LV). B1ϩ profiles were analyzed to determine the amount of variation and dominant patterns of variation across the LV. A total of five to 10 measurements were obtained in each volunteer to determine an upper bound of measurement repeatability. Results:The amount of flip angle variation was found to be 23% to 48% over the LV in mid-short-axis slices and 32% to 63% over the entire LV volume. The standard deviation (SD) of multiple flip angle measurements was Ͻ1.4°over the LV in all subjects, indicating excellent repeatability of the proposed measurement method. The pattern of in-plane flip angle variation was found to be primarily unidirectional across the LV, with a residual variation of Յ3% in all subjects. Conclusion:The in-plane B1ϩ variation over the LV at 3T with body-coil transmission is on the order of 32% to 63% and is predominantly unidirectional in short-axis slices. Reproducible B1ϩ measurements over the whole heart can be obtained in a single breathhold of 16 heartbeats. AS CARDIAC MRI moves to higher field strengths such as 3T, imaging protocols require careful consideration of possible nonuniformity of the transmitted radio frequency (RF) (B1ϩ) field. Knowledge of this nonuniformity is crucial for pulse sequence calibration, imagebased quantitation (such as in first-pass myocardial perfusion imaging), signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) optimization, and the design of new pulse sequences. Estimated variations of 30% to 50% in the B1ϩ field over the heart at 3T have been reported in the literature (1-3). However, the analysis of in vivo B1ϩ variations in the chest has been limited by the lack of time-efficient B1ϩ mapping techniques.There are several existing methods for B1ϩ mapping in static body regions (2-8). One of the simplest and the most straightforward methods is the double angle method (DAM) (6,7), which involves acquiring images with two nominal flip angles (operator prescribed values) ␣ and 2␣. The method uses the trigonometric double angle formula to determine the true flip angle and requires a long repetition time (TR Ͼ Ͼ T1) to ensure full relaxation before ␣ and 2␣ pulses.Cunningham et al (3) recently proposed the saturated double-angle method (SDAM), which permits rapid B1 mapping with TR Ͻ T1. A saturation pulse at the end of each data acquisition resets the longitudinal magnetization to a known state. The B1ϩ field is still derived from the ratio of signal magnitudes after ␣ an...

Purpose: The diagnostic gold standard for nonalcoholic fatty liver disease is an invasive biopsy. Noninvasive Cartesian MRI fat quantification remains limited to a breath-hold (BH). In this work, a novel free-breathing 3D stack-of-radial (FB radial) liver fat quantification technique is developed and evaluated in a preliminary study. Methods: Phantoms and healthy subjects (n ¼ 11) were imaged at 3 Tesla. The proton-density fat fraction (PDFF) determined using FB radial (with and without scan acceleration) was compared to BH single-voxel MR spectroscopy (SVS) and BH 3D Cartesian MRI using linear regression (correlation coefficient r and concordance coefficient r c ) and BlandAltman analysis. Results: In phantoms, PDFF showed significant correlation (r > 0.998, r c > 0.995) and absolute mean differences < 2.2% between FB radial and BH SVS, as well as significant correlation (r > 0.999, r c > 0.998) and absolute mean differences < 0.6% between FB radial and BH Cartesian. In the liver and abdomen, PDFF showed significant correlation (r > 0.986, r c > 0.985) and absolute mean differences < 1% between FB radial and BH SVS, as well as significant correlation (r > 0.996, r c > 0.995) and absolute mean differences < 0.9% between FB radial and BH Cartesian.Conclusion: Accurate 3D liver fat quantification can be performed in 1 to 2 min using a novel FB radial technique. Magn Reson Med 79:370-382,

Purpose To quantify B1+ variation across the breasts and to evaluate the accuracy of pre-contrast T1 estimation with and without B1+ variation in breast MRI patients at 3T. Materials and Methods B1+ and variable flip angle (VFA) T1 mapping were included in our dynamic contrast-enhanced (DCE) breast imaging protocol to study a total of 25 patients on a 3.0T GE MR 750 system. We computed pre-contrast T1 relaxation in fat, which we assumed to be consistent across a cohort of breast imaging subjects, with and without compensation for B1+ variation. The mean and standard deviation of B1+ and T1 values were calculated for statistical data analysis. Results Our measurements showed a consistent B1+ field difference between the left and right breasts. The left breast has an average 15.4% higher flip angle than the prescribed flip angle, and the right breast has an average 17.6% lower flip angle than the prescribed flip angle. This average 33% flip angle difference, which can be vendor and model specific, creates a 52% T1 estimation bias in fat between breasts using the VFA T1 mapping technique. The T1 variation is reduced to 7% by including B1+ correction. Conclusion We have shown that severe B1+ variation over the breasts can cause a substantial error in T1 estimation between the breasts, in VFA T1 maps at 3T, but that compensating for these variations can considerably improve accuracy of T1 measurements, which can directly benefit quantitative breast DCE-MRI at 3T.

GENITOURINARY IMAGINGM ultiparametric MRI is an important tool in the diagnosis of prostate cancer (PCa) (1,2). However, multiparametric MRI still misses PCa in up to 45% of men and faces challenges in distinguishing clinically significant PCa from indolent PCa (2,3). Thus, histopathologic examination of PCa remains the reference standard. A Gleason score based on the microscopic appearance of PCa is assigned to indicate its aggressiveness (4).Diffusion-weighted MRI is a critical component of multiparametric MRI and is sensitive to tissue microstructure changes in PCa (5). However, current clinical analysis using a monoexponential signal model to calculate apparent diffu-Materials and Methods: Men with PCa who underwent 3-T MRI and robotic-assisted radical prostatectomy between June 2018 and January 2019 were prospectively studied. After prostatectomy, the fresh whole prostate specimens were imaged in patient-specific threedimensionally printed molds by using 3-T MRI with DR-CSI and were then sliced to create coregistered WMHP slides. The DR-CSI spectral signal component fractions (f A , f B , f C ) were compared with epithelial, stromal, and luminal area fractions (f epithelium , f stroma , f lumen ) quantified in PCa and benign tissue regions. A linear mixed-effects model assessed the correlations between (f A , f B , f C ) and (f epithelium , f stroma , f lumen ), and the strength of correlations was evaluated by using Spearman correlation coefficients. Differences between PCa and benign tissues in terms of DR-CSI signal components and microscopic tissue compartments were assessed using two-sided t tests.Results: Prostate specimens from nine men (mean age, 65 years 6 7 [standard deviation]) were evaluated; 20 regions from 17 PCas, along with 20 benign tissue regions of interest, were analyzed. Three DR-CSI spectral signal components (spectral peaks) were consistently identified. The f A , f B , and f C were correlated with f epithelium , f stroma , and f lumen (all P , .001), with Spearman correlation coefficients of 0.74 (95% confidence interval [CI]: 0.62, 0.83), 0.80 (95% CI: 0.66, 0.89), and 0.67 (95% CI: 0.51, 0.81), respectively. PCa exhibited differences compared with benign tissues in terms of increased f A (PCa vs benign, 0.37 6 0.05 vs 0.27 6 0.06; P , .001), decreased f C (PCa vs benign, 0.18 6 0.06 vs 0.31 6 0.13; P = .01), increased f epithelium (PCa vs benign, 0.44 6 0.13 vs 0.26 6 0.16; P , .001), and decreased f lumen (PCa vs benign, 0.14 6 0.08 vs 0.27 6 0.18; P = .004). Conclusion:Diffusion-relaxation correlation spectrum imaging signal components correlate with microscopic tissue compartments in the prostate and differ between cancer and benign tissue.