ObjectiveThis study aims to assess the statistical significance of training parameters in 240 dense UNets (DUNets) used for enhancing low Signal-to-Noise Ratio (SNR) and undersampled MRI in various... Show moreObjectiveThis study aims to assess the statistical significance of training parameters in 240 dense UNets (DUNets) used for enhancing low Signal-to-Noise Ratio (SNR) and undersampled MRI in various acquisition protocols. The objective is to determine the validity of differences between different DUNet configurations and their impact on image quality metrics.Materials and MethodsTo achieve this, we trained all DUNets using the same learning rate and number of epochs, with variations in 5 acquisition protocols, 24 loss function weightings, and 2 ground truths. We calculated evaluation metrics for two metric regions of interest (ROI). We employed both Analysis of Variance (ANOVA) and Mixed Effects Model (MEM) to assess the statistical significance of the independent parameters, aiming to compare their efficacy in revealing differences and interactions among fixed parameters.ResultsANOVA analysis showed that, except for the acquisition protocol, fixed variables were statistically insignificant. In contrast, MEM analysis revealed that all fixed parameters and their interactions held statistical significance. This emphasizes the need for advanced statistical analysis in comparative studies, where MEM can uncover finer distinctions often overlooked by ANOVA.DiscussionThese findings highlight the importance of utilizing appropriate statistical analysis when comparing different deep learning models. Additionally, the surprising effectiveness of the UNet architecture in enhancing various acquisition protocols underscores the potential for developing improved methods for characterizing and training deep learning models. This study serves as a stepping stone toward enhancing the transparency and comparability of deep learning techniques for medical imaging applications. Show less
Background: Vaccine hesitancy and lack of access remain major issues in disseminating COVID-19 vaccination to liver patients globally. Factors predicting poor response to vaccination and risk of... Show moreBackground: Vaccine hesitancy and lack of access remain major issues in disseminating COVID-19 vaccination to liver patients globally. Factors predicting poor response to vaccination and risk of breakthrough infection are important data to target booster vaccine programs. The primary aim of the current study was to measure humoral responses to 2 doses of COVID-19 vaccine. Secondary aims included the determination of factors predicting breakthrough infection.Methods: COVID-19 vaccination and Biomarkers in cirrhosis And post-Liver Transplantation is a prospective, multicenter, observational case-control study. Participants were recruited at 4–10 weeks following first and second vaccine doses in cirrhosis [n = 325; 94% messenger RNA (mRNA) and 6% viral vaccine], autoimmune liver disease (AILD) (n = 120; 77% mRNA and 23% viral vaccine), post-liver transplant (LT) (n = 146; 96% mRNA and 3% viral vaccine), and healthy controls (n = 51; 72% mRNA, 24% viral and 4% heterologous combination). Serological end points were measured, and data regarding breakthrough SARS-CoV-2 infection were collected.Results: After adjusting by age, sex, and time of sample collection, anti-Spike IgG levels were the lowest in post-LT patients compared to cirrhosis (p < 0.0001), AILD (p < 0.0001), and control (p = 0.002). Factors predicting reduced responses included older age, Child-Turcotte-Pugh B/C, and elevated IL-6 in cirrhosis; non-mRNA vaccine in AILD; and coronary artery disease, use of mycophenolate and dysregulated B-call activating factor, and lymphotoxin-α levels in LT. Incident infection occurred in 6.6%, 10.6%, 7.4%, and 15.6% of cirrhosis, AILD, post-LT, and control, respectively. The only independent factor predicting infection in cirrhosis was low albumin level.Conclusions: LT patients present the lowest response to the SARS-CoV-2 vaccine. In cirrhosis, the reduced response is associated with older age, stage of liver disease and systemic inflammation, and breakthrough infection with low albumin level. Show less
Sharma, R.; Patelli, A.S.; Bruin, L. de; Maddocks, J.H. 2023
MR scans of low-gamma X-nuclei, low-concentration metabolites, or standard imaging at very low field entail a challenging tradeoff between resolution, signal-to-noise, and acquisition duration.... Show moreMR scans of low-gamma X-nuclei, low-concentration metabolites, or standard imaging at very low field entail a challenging tradeoff between resolution, signal-to-noise, and acquisition duration. Deep learning (DL) techniques, such as UNets, can potentially be used to improve such "low-quality" (LQ) images. We investigate three UNets for upscaling LQ MRI: dense (DUNet), robust (RUNet), and anisotropic (AUNet). These were evaluated for two acquisition scenarios. In the same-subject High-Quality Complementary Priors (HQCP) scenario, an LQ and a high quality (HQ) image are collected and both LQ and HQ were inputs to the UNets. In the No Complementary Priors (NoCP) scenario, only the LQ images are collected and used as the sole input to the UNets. To address the lack of same-subject LQ and HQ images, we added data from the OASIS-1 database. The UNets were tested in upscaling 1/8, 1/4, and 1/2 undersampled images for both scenarios. As manifested by non-statically significant differences of matrices, also supported by subjective observation, the three UNets upscaled images equally well. This was in contrast to mixed effects statistics that clearly illustrated significant differences. Observations suggest that the detailed architecture of these UNets may not play a critical role. As expected, HQCP substantially improves upscaling with any of the UNets. The outcomes support the notion that DL methods may have merit as an integral part of integrated holistic approaches in advancing special MRI acquisitions; however, primary attention should be paid to the foundational step of such approaches, i.e., the actual data collected. Show less
OBJECTIVES The aim of this study was to evaluate the predictors of left ventricular outflow tract (LVOT) obstruction after transcatheter mitral valve replacement (TMVR).BACKGROUND LVOT obstruction... Show moreOBJECTIVES The aim of this study was to evaluate the predictors of left ventricular outflow tract (LVOT) obstruction after transcatheter mitral valve replacement (TMVR).BACKGROUND LVOT obstruction is a major concern with TMVR, but limited data exist regarding its predictors and impact on outcomes.METHODS Patients with pre-procedural multidetector row computed tomography (MDCT) undergoing TMVR for failed mitral bioprosthetic valves (valve-in-valve), annuloplasty rings (valve-in-ring), and mitral annular calcification (valve-in-MAC) were included in this study. Echocardiographic and procedural characteristics were recorded, and comprehensive assessment with MDCT was performed to identify the predictors of LVOT obstruction (defined as an increment of mean LVOT gradient >= 10 mm Hg from baseline). The new LVOT (neo-LVOT) area left after TMVR was estimated by embedding a virtual valve into the mitral annulus on MDCT, simulating the procedure.RESULTS Among 194 patients with pre-procedural MDCT undergoing TMVR (valve-in-valve, 107 patients; valve-in-ring, 50 patients; valve-in-MAC, 37 patients), LVOT obstruction was observed in 26 patients (13.4%), with a higher rate after valve-in-MAC than valve-in-ring and valve-in-valve (54.1% vs. 8.0% vs. 1.9%; p < 0.001). Patients with LVOT obstruction had significantly higher procedural mortality compared with those without LVOT obstruction (34.6% vs. 2.4%; p < 0.001). Receiver-operating characteristic curve analysis showed that an estimated neo-LVOT area <= 1.7 cm(2) predicted LVOT obstruction with sensitivity of 96.2% and specificity of 92.3%.CONCLUSIONS LVOT obstruction after TMVR was associated with higher procedural mortality. A small estimated neo-LVOT area was significantly associated with LVOT obstruction after TMVR and may help identify patients at high risk for LVOT obstruction. (c) 2019 by the American College of Cardiology Foundation. Show less