Explore the vast range of titles available.

Quantification of myocardial scar from late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) images can be facilitated by automated artificial intelligence (AI)-based analysis. However, AI models are susceptible to domain shifts in which the model performance is degraded when applied to data with different characteristics than the original training data. In this study, CycleGAN models were trained to translate local hospital data to the appearance of a public LGE CMR dataset. After domain adaptation, an AI scar quantification pipeline including myocardium segmentation, scar segmentation, and computation of scar burden, previously developed on the public dataset, was evaluated on an external test set including 44 patients clinically assessed for ischemic scar. The mean ± standard deviation Dice similarity coefficients between the manual and AI-predicted segmentations in all patients were similar to those previously reported: 0.76 ± 0.05 for myocardium and 0.75 ± 0.32 for scar, 0.41 ± 0.12 for scar in scans with pathological findings. Bland-Altman analysis showed a mean bias in scar burden percentage of -0.62% with limits of agreement from -8.4% to 7.17%. These results show the feasibility of deploying AI models, trained with public data, for LGE CMR quantification on local clinical data using unsupervised CycleGAN-based domain adaptation. Relevance statement Our study demonstrated the possibility of using AI models trained from public databases to be applied to patient data acquired at a specific institution with different acquisition settings, without additional manual labor to obtain further training labels. Graphical Abstract

BackgroundAssessment of myocardial ischaemia from stress perfusion cardiovascular magnetic resonance (SP-CMR) is largely dependent on visual interpretation, with high level operator experience necessary for accurate assessment. Automated SP-CMR perfusion mapping techniques can provide estimation of myocardial blood flow (MBF). This study investigated the use of novel high spatial-resolution free-breathing SP-CMR with automated quantitative mapping in the diagnosis of coronary artery disease (CAD). Diagnostic performance was evaluated against invasive coronary angiography (ICA) with fractional flow reserve (FFR) measurement.MethodsSeven-hundred and three patients were recruited for SP-CMR using the research sequence at 3 Tesla. Of those receiving ICA within 6 months, 60 patients had FFR measurement. Those with FFR value ≤ 0.80 were classified as having functionally significant CAD. All patients were imaged during vasodilatory hyperaemia in 3 ventricular slices (basal, mid, apical) with dual-sequence acquisition to calculate the arterial input function. Myocardial blood flow (MBF) maps were automatically generated in-line on the scanner following image acquisition at hyperaemic stress and rest, allowing myocardial perfusion reserve (MPR) calculation. MBF maps were analysed using 600 data points per slice, allowing analysis of the endocardial (inner 50%) and epicardial (outer 50%) myocardial layers. Non-parametric statistical analysis was performed using Mann-Whitney U testing and ROC analysis.ResultsSeventy-five coronary vessels assessed by FFR were evaluated at both segmental and coronary territory level. The baseline demographics are shown in table 1. Coronary territory stress MBF and MPR were reduced in FFR-positive (≤ 0.80) regions (median stress MBF: 1.74 [0.90–2.17] ml/min/g; MPR: 1.67 [1.10–1.89]) compared with FFR-negative (> 0.80) regions (stress MBF: 2.50 [2.15–2.95] ml/min/g; MPR 2.35 [2.06–2.54] p < 0.001 for both) [figure 1]. When only analysing the lowest 2 segments of each coronary territory, perfusion values were lower in FFR-positive territories (FFR ≤ 0.80: stress MBF 1.53 [0.79–1.90] ml/min/g, MPR 1.42 [0.99–1.78]; FFR > 0.80: stress MBF 2.08 [1.87–2.52] ml/min/g, MPR 2.11 [1.86–2.45], p < 0.001 for both). A similar pattern is seen when assessing the endocardial layers of each coronary territory (FFR ≤ 0.80: stress MBF 1.63 [0.86–2.03] ml/min/g, MPR 1.63 [1.08–1.87]; FFR > 0.80: stress MBF 2.32 [2.00–2.92] ml/min/g, MPR 2.14 [1.92–2.49], p < 0.001 for both).Considering whole territory analysis, stress MBF ≤ 1.94 ml/min/g and MPR ≤ 1.97 accurately detected FFR-positive CAD on a per-vessel basis (area under the curve: 0.85 and 0.96 respectively; p < 0.001 for both). For whole territory analysis: stress MBF sensitivity 70.0%, specificity 90.8%; MPR sensitivity 90.0%, specificity 84.9%. Similar diagnostic performance is seen for both ‘lowest 2 segments’ and ‘endocardial’ assessment methods, with higher sensitivity and lower specificity seen for both alternative analyses [figure 2].ConclusionScanner-integrated high-resolution free-breathing SP-CMR with automated in-line perfusion mapping can accurately detect functionally significant CAD. High-resolution MBF maps allow assessment of both transmural myocardial segments and the endocardial layers within each coronary territory.Abstract 177 Table 1Baseline characteristics at the time of CMR scan. PCI = Percutaneous coronary intervention; CABG = Coronary artery bypass grafting. Continuous variables displayed as group mean ± standard deviation; discrete variables displayed as frequency (group%). Age 61.3 ± 10.5 Sex Male 31 (51.7) Stress agent Adenosine 56 (93.3) Left ventricular ejection fraction 58.3 ± 9.1 Cardiac rhythm Sinus rhythm 58 (96.7) Atrial fibrillation 2 (3.3) History of CAD 24 (40.0) Previous coronary intervention None 48 (80.0) PCI 11 (18.3) CABG 1 (1.7) Diabetes mellitus 19 (31.7) Hypercholesterolemia 40 (66.7) Hypertension 34 (56.7) Smoking history None 36 (60.0) Ex-smoker 20 (33.3) Current smoker 4 (6.7) Presence of LGE (any myocardial segment) 22 (36.7) Abstract 177 Figure 1Box plots highlighting the differences in both stress MBF (A) and MPR (B) values seen between FFR-negative (> 0.80) and FFR-positive (≤ 0.80) lesions. Similar differences are seen for whole territory, lowest 2 segments and endocardial assessment strategiesAbstract 177 Figure 2ROC curves demonstrating diagnostic performance of whole territory (A), lowest 2 segment (B) and endocardial (C) analysis in the assessment of functionally significant coronary disease as defined by FFRConflict of InterestNone

Written by Thucydides around 400 AD,

Objectives To develop and share a deep learning method that can accurately identify optimal inversion time (TI) from multi-vendor, multi-institutional and multi-field strength inversion scout (TI scout) sequences for late gadolinium enhancement cardiac MRI. Materials and methods Retrospective multicentre study conducted on 1136 1.5-T and 3-T cardiac MRI examinations from four centres and three scanner vendors. Deep learning models, comprising a convolutional neural network (CNN) that provides input to a long short-term memory (LSTM) network, were trained on TI scout pixel data from centres 1 to 3 to identify optimal TI, using ground truth annotations by two readers. Accuracy within 50 ms, mean absolute error (MAE), Lin’s concordance coefficient (LCCC) and reduced major axis regression (RMAR) were used to select the best model from validation results, and applied to holdout test data. Robustness of the best-performing model was also tested on imaging data from centre 4. Results The best model (SE-ResNet18-LSTM) produced accuracy of 96.1%, MAE 22.9 ms and LCCC 0.47 compared to ground truth on the holdout test set and accuracy of 97.3%, MAE 15.2 ms and LCCC 0.64 when tested on unseen external (centre 4) data. Differences in vendor performance were observed, with greatest accuracy for the most commonly represented vendor in the training data. Conclusion A deep learning model was developed that can identify optimal inversion time from TI scout images on multi-vendor data with high accuracy, including on previously unseen external data. We make this model available to the scientific community for further assessment or development. Clinical relevance statement A robust automated inversion time selection tool for late gadolinium–enhanced imaging allows for reproducible and efficient cross-vendor inversion time selection. Key Points • A model comprising convolutional and recurrent neural networks was developed to extract optimal TI from TI scout images. • Model accuracy within 50 ms of ground truth on multi-vendor holdout and external data of 96.1% and 97.3% respectively was achieved. • This model could improve workflow efficiency and standardise optimal TI selection for consistent LGE imaging.

BackgroundBased on the PARADIGM-HF study trial, sacubitril/valsartan (SV) was approved by NICE in April 2016 (TA388) for patients with symptomatic heart failure. SV is recommended in patients with a left ventricular ejection fraction (LVEF) 35% despite a stable dose of ACE inhibitor (ACEi) or angiotensin receptor blocker (ARB), and large numbers are potentially eligible for this first-in-class drug. However, there is a lack of real world experience of both drug tolerability and systems for initiation and monitoring in overstretched heart failure services.MethodsSuitable patients were identified and started on SV by heart failure specialists. Dedicated, registrar-delivered monitoring and up titration clinics were established. Patients were reviewed 2–4 weekly. Symptoms, vital signs, biochemistry and hospital admissions were recorded at each visit. Once stable on optimal doses, patients were discharged to primary care, as pre-arranged with the District Prescribing Committee. Our initial 6 month experience has been analysed.Results69 patients (mean age 63.2±11.6 years) were commenced on SV. Mean LVEF 27.5±6.7%; mean baseline eGFR 66.1±21.9 ml/min/1.73m2. Prior to initiation of SV, mean baseline ACEi/ARB dose was equivalent to 16.3±6.7 mg enalapril daily. Overall 68/69 (98.6%) prescriptions of SV were NICE TA388 compliant (1 patient ACEi/ARB intolerant).9 patients (13.0%) stopped the medication due to adverse effects (PARADIGM-HF 17.8%), whilst another 3 patients (4.3%) were down titrated to a tolerable lower dose. 15.9% of all patients experienced symptomatic hypotension (PARADIGM-HF 14.0%). No episodes of angioedema, nor significant deterioration in renal function (50% reduction in eGFR) were observed. Only 1 (1.4%) patient was hospitalised with decompensated heart failure symptoms, but 3 (4.3%) patients were admitted with syncope secondary to orthostatic hypotension.A total of 36 patients were discharged, with a median ?follow up time of 39 days (IQR 23) from commencement to?stable discharge dose each requiring 1 initiation consultation and a mean of 2.4±1.0 follow up consultations. The majority of patients 25 (69.4%) were discharged at the highest ?dose?–?97/?103 mg BD. 23 (63.9%) of those discharged reported a subjective improvement in symptoms and quality of life.ConclusionsInitiation of SV and dose optimisation in clinical practice represents a significant burden of additional work for heart failure teams. Dedicated, registrar-led outpatient clinics to monitor patients commenced on SV by heart failure specialists can successfully address this.Prescribing within NICE TA388 guidelines in real world patients, there were similar drug tolerance and adverse event rates to those reported in PARADIGM-HF. However, the lower mean age within this particular population, who were carefully selected, may indicate that such findings are not representative of the entire heart failure population.

Background: Quantitative stress perfusion cardiovascular magnetic resonance (CMR) is a powerful tool for assessing myocardial ischemia. Motion correction is essential for accurate pixel-wise mapping but traditional registration-based methods are slow and sensitive to acquisition variability, limiting robustness and scalability. Methods: We developed an unsupervised deep learning-based motion correction pipeline that replaces iterative registration with efficient one-shot estimation. The method corrects motion in three steps and uses robust principal component analysis to reduce contrast-related effects. It aligns the perfusion series and auxiliary images (arterial input function and proton density-weighted series). Models were trained and validated on multivendor data from 201 patients, with 38 held out for testing. Performance was assessed via temporal alignment and quantitative perfusion values, compared to a previously published registration-based method. Results: The deep learning approach significantly improved temporal smoothness of time-intensity curves (p<0.001). Myocardial alignment (Dice = 0.92 (0.04) and 0.91 (0.05)) was comparable to the baseline and superior to before registration (Dice = 0.80 (0.09), p<0.001). Perfusion maps showed reduced motion, with lower standard deviation in the myocardium (0.52 (0.39) ml/min/g) compared to baseline (0.55 (0.44) ml/min/g). Processing time was reduced 15-fold. Conclusion: This deep learning pipeline enables fast, robust motion correction for stress perfusion CMR, improving accuracy across dynamic and auxiliary images. Trained on multivendor data, it generalizes across sequences and may facilitate broader clinical adoption of quantitative perfusion imaging.

Now that Sark has an all-elected parliament (Barclay brothers tell Sark: you didn't...

Planning is a council activity that cannot 'shut down' during electoral purdah restrictions, but extra care should be taken during this time of heightened sensitivity. Planning committee meetings should continue, although councils should think about whether the reporting of outcomes might reflect on particular groups or even individual councillors. The...

By the 1950s most of the uranium districts of the U.S.A. had been discovered as a result of incentives from the U.S. Atomic Energy Commission, and 94% of resources proved to be sandstone deposits. Principal resource regions are the Colorado Plateau, producing about 65% of the U.S. total, Wyoming basins (24%) and the South Texas coastal plain (only 5%). The two major types of deposit found in these regions are 'roll-type' and 'peneconcordant', the former precipitated at geochemical fronts where oxidizing uranium-bearing groundwater penetrated reduced sandstones. The latter occur where uranium in solution has been precipitated locally by agents such as carbonaceous materials, humates or pyrite. Colorado Plateau production is mainly from peneconcordant deposits, the majority from the Grants uranium region (New Mexico) Uravan mineral belt (Colorado) and Lisbon Valley (Utah). These are commonly underground mines, although the largest uranium open pit in the U.S.A. is found in the Grants region. Roll-type deposits of Wyoming, where the Powder River Basin was the most active area of exploitation in the 1970s, are generally mined open pit, though some are underground and have been tested for in-situ leach mining. Both open pit and leach mining are common in Texas. Over 80% of uranium exploration in the U.S.A. since 1974 has been in sandstone deposits, and projections predict that domestic demand to 2000 will be satisfied from U.S. proven reserves and probable potential resources. To meet these requirements, however, most production will be from resources in forward cost categories of more than U.S.$30 per pound U₃O₈. At present U.S. exploration and production have been curtailed, because of over supply, a trend that will continue for some years. Although demand is expected to grow after 1986, foreign low-cost non-sandstone uranium will compete with U.S. reserves, which may already have been adversely affected by mine closures, 'high grading' and decreased exploration and development.