To tune and test the generalizability of a deep learning-based model for assessment of COVID-19 lung disease severity on chest radiographs (CXRs) from different patient populations. A published ...convolutional Siamese neural network-based model previously trained on hospitalized patients with COVID-19 was tuned using 250 outpatient CXRs. This model produces a quantitative measure of COVID-19 lung disease severity (pulmonary x-ray severity (PXS) score). The model was evaluated on CXRs from 4 test sets, including 3 from the United States (patients hospitalized at an academic medical center (N = 154), patients hospitalized at a community hospital (N = 113), and outpatients (N = 108)) and 1 from Brazil (patients at an academic medical center emergency department (N = 303)). Radiologists from both countries independently assigned reference standard CXR severity scores, which were correlated with the PXS scores as a measure of model performance (Pearson R). The Uniform Manifold Approximation and Projection (UMAP) technique was used to visualize the neural network results. Tuning the deep learning model with outpatient data showed high model performance in 2 United States hospitalized patient datasets (R = 0.88 and R = 0.90, compared to baseline R = 0.86). Model performance was similar, though slightly lower, when tested on the United States outpatient and Brazil emergency department datasets (R = 0.86 and R = 0.85, respectively). UMAP showed that the model learned disease severity information that generalized across test sets. A deep learning model that extracts a COVID-19 severity score on CXRs showed generalizable performance across multiple populations from 2 continents, including outpatients and hospitalized patients.
Ultrasound (US) is the first-line imaging tool for evaluating liver and kidney transplants during and after the surgical procedures. In most patients after organ transplantation, gray-scale US ...coupled with color/power and spectral Doppler techniques is used to evaluate the transplant organs, assess the patency of vascular structures, and identify potential complications. In technically difficult or inconclusive cases, however, contrast-enhanced ultrasound (CEUS) can provide prompt and accurate diagnostic information that is essential for management decisions. CEUS is indicated to evaluate for vascular complications including vascular stenosis or thrombosis, active bleeding, pseudoaneurysms and arteriovenous fistulas. Parenchymal indications for CEUS include evaluation for perfusion defects and focal inflammatory and non-inflammatory lesions. When transplant rejection is suspected, CEUS can assist with prompt intervention by excluding potential underlying causes for organ dysfunction. Intracavitary CEUS applications can evaluate the biliary tract of a liver transplant (e.g., for biliary strictures, bile leak or intraductal stones) or the urinary tract of a renal transplant (e.g., for urinary obstruction, urine leak or vesicoureteral reflux) as well as the position and patency of hepatic, biliary and renal drains and catheters. The aim of this review is to present current experience regarding the use of CEUS to evaluate liver and renal transplants, focusing on the examination technique and interpretation of the main imaging findings, predominantly those related to vascular complications.
Contrast-enhanced ultrasound (CEUS) has emerged as a valuable modality for bowel imaging in adults and children. CEUS enables visualization of the perfusion of the bowel wall and of the associated ...mesentery in healthy and disease states. In addition, CEUS images can be used to make quantitative measurements of contrast kinetics, allowing for objective assessment of bowel wall enhancement. Bowel CEUS is commonly applied to evaluate inflammatory bowel disease and to monitor treatment response. It has also been applied to evaluate necrotizing enterocolitis, intussusception, appendicitis and epiploic appendagitis, although experience with these applications is more limited. In this review article, we present the current experience using CEUS to evaluate the pediatric bowel with emphasis on inflammatory bowel disease, extrapolating the established experience from adult studies. We also discuss emerging applications of CEUS as an adjunct or problem-solving tool for evaluating bowel perfusion.
We aimed to characterize the fetal buccal fat pad (BFP) on magnetic resonance imaging (MRI) to determine the frequency and types of sequences on which the BFP demonstrates low signal intensity and ...determine any possible correlation with timing of the MRI during fetal development.
A retrospective review of all fetal MR studies was performed, and a pediatric radiologist blinded to the referring and final fetal diagnosis as well as outcome evaluated the included cases. A positive buccal fat pad sign (BFS) was recorded as present if a round, symmetric, and bilateral area was seen in the submalar region of the face with the following signal characteristics: T1 hyperintensity, low signal on echo planar imaging (EPI), low signal on true fast imaging with steady-state free precession (TRUFI), and with restriction on diffusion-weighted imaging (DWI).
A total of one hundred sixty-seven (167) fetal MRI studies: one hundred fourteen (114) body (68%) and fifty-three (53) neuro (32%) scans were reviewed during the study period. The BFS was most commonly seen on EPI (63%) and TRUFI (49%) sequences. Substantial agreement between TRUFI and EPI (κ = 0.68;
< 0.01); moderate agreement between TRUFI and T1 (κ = 0.53;
< 0.01) as well as T1 and EPI (κ = 0.53;
< 0.01), and fair agreement between EPI and Diffusion (κ = 0.28;
< 0.01) was observed. The median gestational age (GA) was 24 weeks (IQR 22-30 weeks). The fetuses with a positive BFS were significantly older (mean GA of 27 weeks or higher) than those without, for each sequence.
The focal low signal in the fetal buccal fat pad, termed the fetal BFS, is a commonly encountered normal finding in the majority of fetal MRI scans on TRUFI and EPI sequences. This finding may be related to the presence and development of brown adipose tissue in the buccal fat pad resulting in T2* effects, but further studies are needed in order to confirm this. Further work can incorporate any of the sensitive sequences demonstrating low signal in brown adipose tissue to map its distribution and development in the fetus and beyond.
Rotenone exposure in rodents provides an interesting model for studying mechanisms of toxin-induced dopaminergic neuronal injury. However, several aspects remain unclear regarding the effects and the ...accuracy of rotenone as an animal model of Parkinson's disease (PD). In this study, we investigated the motor and depressive-like behaviors associated to neurochemical alterations induced by a novel protocol of rotenone administration.
In the first experiment, we adopted the paw test to characterize an effective dose of rotenone able to promote nigrostriatal toxicity. In the second experiment, control and rotenone 2.5 mg/kg groups were injected (ip) for 10 consecutive days.
This test indicated that intraperitonial (ip) rotenone at 2.5 and 5.0 mg/kg promoted a significant neurotoxicity to striatum and nucleus accumbens. However, only 2.5 mg/kg of rotenone was associated to a negligible mortality rate. Open-field tests were conducted on 1, 7, 14 and 21 day after the last day of treatment and showed an important locomotor impairment, confined to 1 and 7 day. Besides, rotenone affected dopamine levels and increased its turnover in the striatum. Modified forced swim test (performed on 22 day) and sucrose preference test (performed on 14 and 21 day) demonstrated that rotenone produced impairments in the swimming and immobility. In parallel, increments in the serotonin and noradrenaline turnovers were observed in the striatum and hippocampus of the rotenone group.
These data suggest important participations of serotonin and noradrenaline in depressive-like behaviors induced by rotenone. Thus, it is proposed that the current rotenone protocol provides an improvement regarding the existing rotenonemodels of PD.
The effect of systemic administration of the cannabinoid antagonist SR 141716A (
N-(piperidin-1-yl)-5-(4-chlorophenyl)-4-methyl-1H-pyrazole-3-carboxyamide) on penile erection and yawning induced by ...apomorphine was investigated in rats. SR 141716A (2 mg/kg, i.p.) administered 40 min before apomorphine (40 and 80 μg/kg, s.c.) increased the number of penile erection and yawning responses. The administration of cannabinoid agonist Δ
9-tetrahydrocannabinol (1.25 mg/kg, i.p.) 15 min before apomorphine (40 and 80 μg/kg, s.c.) did not affect penile erection, however it decreased yawning. The present results provide additional evidence that cannabinoid agonists interfere with dopaminergic systems and that SR 141716A together with a dopaminergic agonist could be useful to potentiate dopaminergic activity.
Stroke is a leading cause of death and disability. The ability to quickly identify the presence of acute infarct and quantify the volume on magnetic resonance imaging (MRI) has important treatment ...implications. We developed a machine learning model that used the apparent diffusion coefficient and diffusion weighted imaging series. It was trained on 6,657 MRI studies from Massachusetts General Hospital (MGH; Boston, USA). All studies were labelled positive or negative for infarct (classification annotation) with 377 having the region of interest outlined (segmentation annotation). The different annotation types facilitated training on more studies while not requiring the extensive time to manually segment every study. We initially validated the model on studies sequestered from the training set. We then tested the model on studies from three clinical scenarios: consecutive stroke team activations for 6-months at MGH, consecutive stroke team activations for 6-months at a hospital that did not provide training data (Brigham and Women's Hospital BWH; Boston, USA), and an international site (Diagnósticos da América SA DASA; Brazil). The model results were compared to radiologist ground truth interpretations. The model performed better when trained on classification and segmentation annotations (area under the receiver operating curve AUROC 0.995 95% CI 0.992-0.998 and median Dice coefficient for segmentation overlap of 0.797 IQR 0.642-0.861) compared to segmentation annotations alone (AUROC 0.982 95% CI 0.972-0.990 and Dice coefficient 0.776 IQR 0.584-0.857). The model accurately identified infarcts for MGH stroke team activations (AUROC 0.964 95% CI 0.943-0.982, 381 studies), BWH stroke team activations (AUROC 0.981 95% CI 0.966-0.993, 247 studies), and at DASA (AUROC 0.998 95% CI 0.993-1.000, 171 studies). The model accurately segmented infarcts with Pearson correlation comparing model output and ground truth volumes between 0.968 and 0.986 for the three scenarios. Acute infarct can be accurately detected and segmented on MRI in real-world clinical scenarios using a machine learning model.