Coronary sinus reducer for the treatment of refractory angina (ORBITA-COSMIC): a randomised, placebo-controlled trial.

BACKGROUND: The coronary sinus reducer (CSR) is proposed to reduce angina in patients with stable coronary artery disease by improving myocardial perfusion. We aimed to measure its efficacy, compared with placebo, on myocardial ischaemia reduction and symptom improvement. METHODS: ORBITA-COSMIC was a double-blind, randomised, placebo-controlled trial conducted at six UK hospitals. Patients aged 18 years or older with angina, stable coronary artery disease, ischaemia, and no further options for treatment were eligible. All patients completed a quantitative adenosine-stress perfusion cardiac magnetic resonance scan, symptom and quality-of-life questionnaires, and a treadmill exercise test before entering a 2-week symptom assessment phase, in which patients reported their angina symptoms using a smartphone application (ORBITA-app). Patients were randomly assigned (1:1) to receive either CSR or placebo. Both participants and investigators were masked to study assignment. After the CSR implantation or placebo procedure, patients entered a 6-month blinded follow-up phase in which they reported their daily symptoms in the ORBITA-app. At 6 months, all assessments were repeated. The primary outcome was myocardial blood flow in segments designated ischaemic at enrolment during the adenosine-stress perfusion cardiac magnetic resonance scan. The primary symptom outcome was the number of daily angina episodes. Analysis was done by intention-to-treat and followed Bayesian methodology. The study is registered with ClinicalTrials.gov, NCT04892537, and completed. FINDINGS: Between May 26, 2021, and June 28, 2023, 61 patients were enrolled, of whom 51 (44 [86%] male; seven [14%] female) were randomly assigned to either the CSR group (n=25) or the placebo group (n=26). Of these, 50 patients were included in the intention-to-treat analysis (24 in the CSR group and 26 in the placebo group). 454 (57%) of 800 imaged cardiac segments were ischaemic at enrolment, with a median stress myocardial blood flow of 1·08 mL/min per g (IQR 0·77-1·41). Myocardial blood flow in ischaemic segments did not improve with CSR compared with placebo (difference 0·06 mL/min per g [95% CrI -0·09 to 0·20]; Pr(Benefit)=78·8%). The number of daily angina episodes was reduced with CSR compared with placebo (OR 1·40 [95% CrI 1·08 to 1·83]; Pr(Benefit)=99·4%). There were two CSR embolisation events in the CSR group, and no acute coronary syndrome events or deaths in either group. INTERPRETATION: ORBITA-COSMIC found no evidence that the CSR improved transmural myocardial perfusion, but the CSR did improve angina compared with placebo. These findings provide evidence for the use of CSR as a further antianginal option for patients with stable coronary artery disease. FUNDING: Medical Research Council, Imperial College Healthcare Charity, National Institute for Health and Care Research Imperial Biomedical Research Centre, St Mary's Coronary Flow Trust, British Heart Foundation.

Quality assurance of late gadolinium enhancement cardiac MRI images: a deep learning classifier for confidence in the presence or absence of abnormality with potential to prompt real-time image optimisation.

BACKGROUND: Late gadolinium enhancement (LGE) of the myocardium has significant diagnostic and prognostic implications, with even small areas of enhancement being important. Distinguishing between definitely normal and definitely abnormal LGE images is usually straightforward; but diagnostic uncertainty arises when reporters are not sure whether the observed LGE is genuine or not. This uncertainty might be resolved by repetition (to remove artefact) or further acquisition of intersecting images, but this must take place before the scan finishes. Real-time quality assurance by humans is a complex task requiring training and experience, so being able to identify which images have an intermediate likelihood of LGE while the scan is ongoing, without the presence of an expert is of high value. This decision-support could prompt immediate image optimisation or acquisition of supplementary images to confirm or refute the presence of genuine LGE. This could reduce ambiguity in reports. METHODS: Short-axis, phase sensitive inversion recovery (PSIR) late gadolinium images were extracted from our clinical CMR database and shuffled. Two, independent, blinded experts scored each individual slice for 'LGE likelihood' on a visual analogue scale, from 0 (absolute certainty of no LGE) to 100 (absolute certainty of LGE), with 50 representing clinical equipoise. The scored images were split into 2 classes - either "high certainty" of whether LGE was present or not, or "low certainty". The dataset was split into training, validation and test sets (70:15:15). A deep learning binary classifier based on the EfficientNetV2 convolutional neural network architecture was trained to distinguish between these categories. Classifier performance on the test set was evaluated by calculating the accuracy, precision, recall, F1-score, and area under the receiver operating characteristics curve (ROC AUC). Performance was also evaluated on an external test set of images from a different centre. RESULTS: 1645 images (from 272 patients) were labelled and split at the patient level into training (1151 images), validation (247 images) and test (247 images) sets for the deep learning binary classifier. Of these, 1208 images were 'high certainty' (255 for LGE, 953 for no LGE), and 437 were 'low certainty'). An external test comprising 247 images from 41 patients from another centre was also employed. After 100 epochs the performance on the internal test set was: accuracy = 94%, recall = 0.80, precision = 0.97, F1-score = 0.87 and ROC AUC = 0.94. The classifier also performed robustly on the external test set (accuracy = 91%, recall = 0.73, precision = 0.93, F1-score = 0.82 and ROC AUC = 0.91). These results were benchmarked against a reference inter-expert accuracy of 86%. CONCLUSIONS: Deep learning shows potential to automate quality control of late gadolinium imaging in CMR. The ability to identify short-axis images with intermediate LGE likelihood in real-time may serve as a useful decision support tool. This approach has the potential to guide immediate further imaging while the patient is still in the scanner, thereby reducing the frequency of recalls and inconclusive reports due to diagnostic indecision.

A double-blind, randomised, placebo-controlled trial of the coronary sinus Reducer in refractory angina: design and rationale of the ORBITA-COSMIC trial.

The coronary sinus Reducer (CSR) is an hourglass-shaped device which creates an artificial stenosis in the coronary sinus. Whilst placebo-controlled data show an improvement in angina, these results are unreplicated and are the subject of further confirmatory research. The mechanism of action of this unintuitive therapy is unknown. The Coronary Sinus Reducer Objective Impact on Symptoms, MRI Ischaemia, and Microvascular Resistance (ORBITA-COSMIC) trial is a randomised, placebo-controlled, double-blind trial investigating the efficacy of the CSR. Patients with (i) established epicardial coronary artery disease, (ii) angina on maximally tolerated antianginal medication, (iii) evidence of myocardial ischaemia and (iv) no further options for percutaneous coronary intervention or coronary artery bypass grafting will be enrolled. Upon enrolment, angina and quality-of-life questionnaires, treadmill exercise testing and quantitative stress perfusion cardiac magnetic resonance (CMR) imaging will be performed. Participants will record their symptoms daily on a smartphone application throughout the trial. After a 2-week symptom assessment phase, participants will be randomised in the cardiac catheterisation laboratory to CSR or a placebo procedure. After 6 months of blinded follow-up, all prerandomisation tests will be repeated. A prespecified subgroup will undergo invasive coronary physiology assessment at prerandomisation and follow-up. The primary outcome is stress myocardial blood flow on CMR. Secondary outcomes include angina frequency, quality of life and treadmill exercise time. (ClinicalTrials.gov: NCT04892537).

Aortic stenosis assessment from the 3-chamber cine: Ratio of balanced steady-state-free-precession (bSSFP) blood signal between the aorta and left ventricle predicts severity.

BACKGROUND: Cardiovascular magnetic resonance (CMR) imaging is an important tool for evaluating the severity of aortic stenosis (AS), co-existing aortic disease, and concurrent myocardial abnormalities. Acquiring this additional information requires protocol adaptations and additional scanner time, but is not necessary for the majority of patients who do not have AS. We observed that the relative signal intensity of blood in the ascending aorta on a balanced steady state free precession (bSSFP) 3-chamber cine was often reduced in those with significant aortic stenosis. We investigated whether this effect could be quantified and used to predict AS severity in comparison to existing gold-standard measurements. METHODS: Multi-centre, multi-vendor retrospective analysis of patients with AS undergoing CMR and transthoracic echocardiography (TTE). Blood signal intensity was measured in a ∼1 cm2 region of interest (ROI) in the aorta and left ventricle (LV) in the 3-chamber bSSFP cine. Because signal intensity varied across patients and scanner vendors, a ratio of the mean signal intensity in the aorta ROI to the LV ROI (Ao:LV) was used. This ratio was compared using Pearson correlations against TTE parameters of AS severity: aortic valve peak velocity, mean pressure gradient and the dimensionless index. The study also assessed whether field strength (1.5 T vs. 3 T) and patient characteristics (presence of bicuspid aortic valves (BAV), dilated aortic root and low flow states) altered this signal relationship. RESULTS: 314 patients (median age 69 [IQR 57-77], 64% male) who had undergone both CMR and TTE were studied; 84 had severe AS, 78 had moderate AS, 66 had mild AS and 86 without AS were studied as a comparator group. The median time between CMR and TTE was 12 weeks (IQR 4-26). The Ao:LV ratio at 1.5 T strongly correlated with peak velocity (r = -0.796, p = 0.001), peak gradient (r = -0.772, p = 0.001) and dimensionless index (r = 0.743, p = 0.001). An Ao:LV ratio of < 0.86 was 84% sensitive and 82% specific for detecting AS of any severity and a ratio of 0.58 was 83% sensitive and 92% specific for severe AS. The ability of Ao:LV ratio to predict AS severity remained for patients with bicuspid aortic valves, dilated aortic root or low indexed stroke volume. The relationship between Ao:LV ratio and AS severity was weaker at 3 T. CONCLUSIONS: The Ao:LV ratio, derived from bSSFP 3-chamber cine images, shows a good correlation with existing measures of AS severity. It demonstrates utility at 1.5 T and offers an easily calculable metric that can be used at the time of scanning or automated to identify on an adaptive basis which patients benefit from dedicated imaging to assess which patients should have additional sequences to assess AS.

A Placebo-Controlled Trial of Percutaneous Coronary Intervention for Stable Angina.

BACKGROUND: Percutaneous coronary intervention (PCI) is frequently performed to reduce the symptoms of stable angina. Whether PCI relieves angina more than a placebo procedure in patients who are not receiving antianginal medication remains unknown. METHODS: We conducted a double-blind, randomized, placebo-controlled trial of PCI in patients with stable angina. Patients stopped all antianginal medications and underwent a 2-week symptom assessment phase before randomization. Patients were then randomly assigned in a 1:1 ratio to undergo PCI or a placebo procedure and were followed for 12 weeks. The primary end point was the angina symptom score, which was calculated daily on the basis of the number of angina episodes that occurred on a given day, the number of antianginal medications prescribed on that day, and clinical events, including the occurrence of unblinding owing to unacceptable angina or acute coronary syndrome or death. Scores range from 0 to 79, with higher scores indicating worse health status with respect to angina. RESULTS: A total of 301 patients underwent randomization: 151 to the PCI group and 150 to the placebo group. The mean (±SD) age was 64±9 years, and 79% were men. Ischemia was present in one cardiac territory in 242 patients (80%), in two territories in 52 patients (17%), and in three territories in 7 patients (2%). In the target vessels, the median fractional flow reserve was 0.63 (interquartile range, 0.49 to 0.75), and the median instantaneous wave-free ratio was 0.78 (interquartile range, 0.55 to 0.87). At the 12-week follow-up, the mean angina symptom score was 2.9 in the PCI group and 5.6 in the placebo group (odds ratio, 2.21; 95% confidence interval, 1.41 to 3.47; P<0.001). One patient in the placebo group had unacceptable angina leading to unblinding. Acute coronary syndromes occurred in 4 patients in the PCI group and in 6 patients in the placebo group. CONCLUSIONS: Among patients with stable angina who were receiving little or no antianginal medication and had objective evidence of ischemia, PCI resulted in a lower angina symptom score than a placebo procedure, indicating a better health status with respect to angina. (Funded by the National Institute for Health and Care Research Imperial Biomedical Research Centre and others; ORBITA-2 ClinicalTrials.gov number, NCT03742050.).

Left bundle branch pacing with and without anodal capture: impact on ventricular activation pattern and acute haemodynamics.

AIMS: Left bundle branch pacing (LBBP) can deliver physiological left ventricular activation, but typically at the cost of delayed right ventricular (RV) activation. Right ventricular activation can be advanced through anodal capture, but there is uncertainty regarding the mechanism by which this is achieved, and it is not known whether this produces haemodynamic benefit. METHODS AND RESULTS: We recruited patients with LBBP leads in whom anodal capture eliminated the terminal R-wave in lead V1. Ventricular activation pattern, timing, and high-precision acute haemodynamic response were studied during LBBP with and without anodal capture. We recruited 21 patients with a mean age of 67 years, of whom 14 were males. We measured electrocardiogram timings and haemodynamics in all patients, and in 16, we also performed non-invasive mapping. Ventricular epicardial propagation maps demonstrated that RV septal myocardial capture, rather than right bundle capture, was the mechanism for earlier RV activation. With anodal capture, QRS duration and total ventricular activation times were shorter (116 ± 12 vs. 129 ± 14 ms, P < 0.01 and 83 ± 18 vs. 90 ± 15 ms, P = 0.01). This required higher outputs (3.6 ± 1.9 vs. 0.6 ± 0.2 V, P < 0.01) but without additional haemodynamic benefit (mean difference -0.2 ± 3.8 mmHg compared with pacing without anodal capture, P = 0.2). CONCLUSION: Left bundle branch pacing with anodal capture advances RV activation by stimulating the RV septal myocardium. However, this requires higher outputs and does not improve acute haemodynamics. Aiming for anodal capture may therefore not be necessary.

Septal scar as a barrier to left bundle branch area pacing.

BACKGROUND: The use of left bundle branch area pacing (LBBAP) for bradycardia pacing and cardiac resynchronization is increasing, but implants are not always successful. We prospectively studied consecutive patients to determine whether septal scar contributes to implant failure. METHODS: Patients scheduled for bradycardia pacing or cardiac resynchronization therapy were prospectively enrolled. Recruited patients underwent preprocedural scar assessment by cardiac MRI with late gadolinium enhancement imaging. LBBAP was attempted using a lumenless lead (Medtronic 3830) via a transeptal approach. RESULTS: Thirty-five patients were recruited: 29 male, mean age 68 years, 10 ischemic, and 16 non-ischemic cardiomyopathy. Pacing indication was bradycardia in 26% and cardiac resynchronization in 74%. The lead was successfully deployed to the left ventricular septum in 30/35 (86%) and unsuccessful in the remaining 5/35 (14%). Septal late gadolinium enhancement was significantly less extensive in patients where left septal lead deployment was successful, compared those where it was unsuccessful (median 8%, IQR 2%-18% vs. median 54%, IQR 53%-57%, p < .001). CONCLUSIONS: The presence of septal scar appears to make it more challenging to deploy a lead to the left ventricular septum via the transeptal route. Additional implant tools or alternative approaches may be required in patients with extensive septal scar.

Efficient labelling for efficient deep learning: the benefit of a multiple-image-ranking method to generate high volume training data applied to ventricular slice level classification in cardiac MRI.

Background: Getting the most value from expert clinicians' limited labelling time is a major challenge for artificial intelligence (AI) development in clinical imaging. We present a novel method for ground-truth labelling of cardiac magnetic resonance imaging (CMR) image data by leveraging multiple clinician experts ranking multiple images on a single ordinal axis, rather than manual labelling of one image at a time. We apply this strategy to train a deep learning (DL) model to classify the anatomical position of CMR images. This allows the automated removal of slices that do not contain the left ventricular (LV) myocardium. Methods: Anonymised LV short-axis slices from 300 random scans (3,552 individual images) were extracted. Each image's anatomical position relative to the LV was labelled using two different strategies performed for 5 hours each: (I) 'one-image-at-a-time': each image labelled according to its position: 'too basal', 'LV', or 'too apical' individually by one of three experts; and (II) 'multiple-image-ranking': three independent experts ordered slices according to their relative position from 'most-basal' to 'most apical' in batches of eight until each image had been viewed at least 3 times. Two convolutional neural networks were trained for a three-way classification task (each model using data from one labelling strategy). The models' performance was evaluated by accuracy, F1-score, and area under the receiver operating characteristics curve (ROC AUC). Results: After excluding images with artefact, 3,323 images were labelled by both strategies. The model trained using labels from the 'multiple-image-ranking strategy' performed better than the model using the 'one-image-at-a-time' labelling strategy (accuracy 86% vs. 72%, P=0.02; F1-score 0.86 vs. 0.75; ROC AUC 0.95 vs. 0.86). For expert clinicians performing this task manually the intra-observer variability was low (Cohen's κ=0.90), but the inter-observer variability was higher (Cohen's κ=0.77). Conclusions: We present proof of concept that, given the same clinician labelling effort, comparing multiple images side-by-side using a 'multiple-image-ranking' strategy achieves ground truth labels for DL more accurately than by classifying images individually. We demonstrate a potential clinical application: the automatic removal of unrequired CMR images. This leads to increased efficiency by focussing human and machine attention on images which are needed to answer clinical questions.

Smartphone-Based Remote Monitoring for Chronic Heart Failure: Mixed Methods Analysis of User Experience From Patient and Nurse Perspectives.

BACKGROUND: Community-based management by heart failure specialist nurses (HFSNs) is key to improving self-care in heart failure with reduced ejection fraction. Remote monitoring (RM) can aid nurse-led management, but in the literature, user feedback evaluation is skewed in favor of the patient rather than nursing user experience. Furthermore, the ways in which different groups use the same RM platform at the same time are rarely directly compared in the literature. We present a balanced semantic analysis of user feedback from patient and nurse perspectives of Luscii, a smartphone-based RM strategy combining self-measurement of vital signs, instant messaging, and e-learning. OBJECTIVE: This study aims to (1) evaluate how patients and nurses use this type of RM (usage type), (2) evaluate patients' and nurses' user feedback on this type of RM (user experience), and (3) directly compare the usage type and user experience of patients and nurses using the same type of RM platform at the same time. METHODS: We performed a retrospective usage type and user experience evaluation of the RM platform from the perspective of both patients with heart failure with reduced ejection fraction and the HFSNs using the platform to manage them. We conducted semantic analysis of written patient feedback provided via the platform and a focus group of 6 HFSNs. Additionally, as an indirect measure of tablet adherence, self-measured vital signs (blood pressure, heart rate, and body mass) were extracted from the RM platform at onboarding and 3 months later. Paired 2-tailed t tests were used to evaluate differences between mean scores across the 2 timepoints. RESULTS: A total of 79 patients (mean age 62 years; 35%, 28/79 female) were included. Semantic analysis of usage type revealed extensive, bidirectional information exchange between patients and HFSNs using the platform. Semantic analysis of user experience demonstrates a range of positive and negative perspectives. Positive impacts included increased patient engagement, convenience for both user groups, and continuity of care. Negative impacts included information overload for patients and increased workload for nurses. After the patients used the platform for 3 months, they showed significant reductions in heart rate (P=.004) and blood pressure (P=.008) but not body mass (P=.97) compared with onboarding. CONCLUSIONS: Smartphone-based RM with messaging and e-learning facilitates bilateral information sharing between patients and nurses on a range of topics. Patient and nurse user experience is largely positive and symmetrical, but there are possible negative impacts on patient attention and nurse workload. We recommend RM providers involve patient and nurse users in platform development, including recognition of RM usage in nursing job plans.

Smartphone-Based Remote Monitoring in Heart Failure With Reduced Ejection Fraction: Retrospective Cohort Study of Secondary Care Use and Costs.

BACKGROUND: Despite effective therapies, the economic burden of heart failure with reduced ejection fraction (HFrEF) is driven by frequent hospitalizations. Treatment optimization and admission avoidance rely on frequent symptom reviews and monitoring of vital signs. Remote monitoring (RM) aims to prevent admissions by facilitating early intervention, but the impact of noninvasive, smartphone-based RM of vital signs on secondary health care use and costs in the months after a new diagnosis of HFrEF is unknown. OBJECTIVE: The purpose of this study is to conduct a secondary care health use and health-economic evaluation for patients with HFrEF using smartphone-based noninvasive RM and compare it with matched controls receiving usual care without RM. METHODS: We conducted a retrospective study of 2 cohorts of newly diagnosed HFrEF patients, matched 1:1 for demographics, socioeconomic status, comorbidities, and HFrEF severity. They are (1) the RM group, with patients using the RM platform for >3 months and (2) the control group, with patients referred before RM was available who received usual heart failure care without RM. Emergency department (ED) attendance, hospital admissions, outpatient use, and the associated costs of this secondary care activity were extracted from the Discover data set for a 3-month period after diagnosis. Platform costs were added for the RM group. Secondary health care use and costs were analyzed using Kaplan-Meier event analysis and Cox proportional hazards modeling. RESULTS: A total of 146 patients (mean age 63 years; 42/146, 29% female) were included (73 in each group). The groups were well-matched for all baseline characteristics except hypertension (P=.03). RM was associated with a lower hazard of ED attendance (hazard ratio [HR] 0.43; P=.02) and unplanned admissions (HR 0.26; P=.02). There were no differences in elective admissions (HR 1.03, P=.96) or outpatient use (HR 1.40; P=.18) between the 2 groups. These differences were sustained by a univariate model controlling for hypertension. Over a 3-month period, secondary health care costs were approximately 4-fold lower in the RM group than the control group, despite the additional cost of RM itself (mean cost per patient GBP £465, US $581 vs GBP £1850, US $2313, respectively; P=.04). CONCLUSIONS: This retrospective cohort study shows that smartphone-based RM of vital signs is feasible for HFrEF. This type of RM was associated with an approximately 2-fold reduction in ED attendance and a 4-fold reduction in emergency admissions over just 3 months after a new diagnosis with HFrEF. Costs were significantly lower in the RM group without increasing outpatient demand. This type of RM could be adjunctive to standard care to reduce admissions, enabling other resources to help patients unable to use RM.

Comparison of methods for delivering cardiac resynchronization therapy: an acute electrical and haemodynamic within-patient comparison of left bundle branch area, His bundle, and biventricular pacing.

AIMS: Left bundle branch area pacing (LBBAP) is a promising method for delivering cardiac resynchronization therapy (CRT), but its relative physiological effectiveness compared with His bundle pacing (HBP) is unknown. We conducted a within-patient comparison of HBP, LBBAP, and biventricular pacing (BVP). METHODS AND RESULTS: Patients referred for CRT were recruited. We assessed electrical response using non-invasive mapping, and acute haemodynamic response using a high-precision haemodynamic protocol. Nineteen patients were recruited: 14 male, mean LVEF of 30%. Twelve had time for BVP measurements. All three modalities reduced total ventricular activation time (TVAT), (ΔTVATHBP -43 ± 14 ms and ΔTVATLBBAP -35 ± 20 ms vs. ΔTVATBVP -19 ± 30 ms, P = 0.03 and P = 0.1, respectively). HBP produced a significantly greater reduction in TVAT compared with LBBAP in all 19 patients (-46 ± 15 ms, -36 ± 17 ms, P = 0.03). His bundle pacing and LBBAP reduced left ventricular activation time (LVAT) more than BVP (ΔLVATHBP -43 ± 16 ms, P < 0.01 vs. BVP, ΔLVATLBBAP -45 ± 17 ms, P < 0.01 vs. BVP, ΔLVATBVP -13 ± 36 ms), with no difference between HBP and LBBAP (P = 0.65). Acute systolic blood pressure was increased by all three modalities. In the 12 with BVP, greater improvement was seen with HBP and LBBAP (6.4 ± 3.8 mmHg BVP, 8.1 ± 3.8 mmHg HBP, P = 0.02 vs. BVP and 8.4 ± 8.2 mmHg for LBBAP, P = 0.3 vs. BVP), with no difference between HBP and LBBAP (P = 0.8). CONCLUSION: HBP delivered better ventricular resynchronization than LBBAP because right ventricular activation was slower during LBBAP. But LBBAP was not inferior to HBP with respect to LV electrical resynchronization and acute haemodynamic response.

Automated Inline Myocardial Segmentation of Joint T1 and T2 Mapping Using Deep Learning.

Purpose: To develop an artificial intelligence (AI) solution for automated segmentation and analysis of joint cardiac MRI short-axis T1 and T2 mapping. Materials and Methods: In this retrospective study, a joint T1 and T2 mapping sequence was used to acquire 4240 maps from 807 patients across two hospitals between March and November 2020. Five hundred nine maps from 94 consecutive patients were assigned to a holdout testing set. A convolutional neural network was trained to segment the endocardial and epicardial contours with use of an edge probability estimation approach. Training labels were segmented by an expert cardiologist. Predicted contours were processed to yield mapping values for each of the 16 American Heart Association segments. Network segmentation performance and segment-wise measurements on the testing set were compared with those of two experts on the holdout testing set. The AI model was fully integrated using open-source software to run on MRI scanners. Results: A total of 3899 maps (92%) were deemed artifact-free and suitable for human segmentation. AI segmentation closely matched that of each expert (mean Dice coefficient, 0.82 ± 0.07 [SD] vs expert 1 and 0.86 ± 0.06 vs expert 2) and compared favorably with interexpert agreement (Dice coefficient, 0.84 ± 0.06 for expert 1 vs expert 2). AI-derived segment-wise values for native T1, postcontrast T1, and T2 mapping correlated with expert-derived values (R 2 = 0.96, 0.98, and 0.87, respectively, vs expert 1, and 0.97, 0.99, and 0.92 vs expert 2) and fell within the range of interexpert reproducibility (R 2 = 0.97, 0.99, and 0.90, respectively). The AI model has since been deployed at two hospitals, enabling automated inline analysis. Conclusion: Automated inline analysis of joint T1 and T2 mapping allows accurate segment-wise tissue characterization, with performance equivalent to that of human experts.Keywords: MRI, Neural Networks, Cardiac, Heart Supplemental material is available for this article. © RSNA, 2022.

Automated multi-beat tissue Doppler echocardiography analysis using deep neural networks.

Tissue Doppler imaging is an essential echocardiographic technique for the non-invasive assessment of myocardial blood velocity. Image acquisition and interpretation are performed by trained operators who visually localise landmarks representing Doppler peak velocities. Current clinical guidelines recommend averaging measurements over several heartbeats. However, this manual process is both time-consuming and disruptive to workflow. An automated system for accurate beat isolation and landmark identification would be highly desirable. A dataset of tissue Doppler images was annotated by three cardiologist experts, providing a gold standard and allowing for observer variability comparisons. Deep neural networks were trained for fully automated predictions on multiple heartbeats and tested on tissue Doppler strips of arbitrary length. Automated measurements of peak Doppler velocities show good Bland-Altman agreement (average standard deviation of 0.40 cm/s) with consensus expert values; less than the inter-observer variability (0.65 cm/s). Performance is akin to individual experts (standard deviation of 0.40 to 0.75 cm/s). Our approach allows for > 26 times as many heartbeats to be analysed, compared to a manual approach. The proposed automated models can accurately and reliably make measurements on tissue Doppler images spanning several heartbeats, with performance indistinguishable from that of human experts, but with significantly shorter processing time. HIGHLIGHTS: • Novel approach successfully identifies heartbeats from Tissue Doppler Images • Accurately measures peak velocities on several heartbeats • Framework is fast and can make predictions on arbitrary length images • Patient dataset and models made public for future benchmark studies.

Automatic Diagnosis Labeling of Cardiovascular MRI by Using Semisupervised Natural Language Processing of Text Reports.

PURPOSE: To assess whether the semisupervised natural language processing (NLP) of text from clinical radiology reports could provide useful automated diagnosis categorization for ground truth labeling to overcome manual labeling bottlenecks in the machine learning pipeline. MATERIALS AND METHODS: In this retrospective study, 1503 text cardiac MRI reports from 2016 to 2019 were manually annotated for five diagnoses by clinicians: normal, dilated cardiomyopathy (DCM), hypertrophic cardiomyopathy, myocardial infarction (MI), and myocarditis. A semisupervised method that uses bidirectional encoder representations from transformers (BERT) pretrained on 1.14 million scientific publications was fine-tuned by using the manually extracted labels, with a report dataset split into groups of 801 for training, 302 for validation, and 400 for testing. The model's performance was compared with two traditional NLP models: a rule-based model and a support vector machine (SVM) model. The models' F1 scores and receiver operating characteristic curves were used to analyze performance. RESULTS: After 15 epochs, the F1 scores on the test set of 400 reports were as follows: normal, 84%; DCM, 79%; hypertrophic cardiomyopathy, 86%; MI, 91%; and myocarditis, 86%. The pooled F1 score and area under the receiver operating curve were 86% and 0.96, respectively. On the same test set, the BERT model had a higher performance than the rule-based model (F1 score, 42%) and SVM model (F1 score, 82%). Diagnosis categories classified by using the BERT model performed the labeling of 1000 MR images in 0.2 second. CONCLUSION: The developed model used labels extracted from radiology reports to provide automated diagnosis categorization of MR images with a high level of performance.Keywords: Semisupervised Learning, Diagnosis/Classification/Application Domain, Named Entity Recognition, MRI Supplemental material is available for this article. © RSNA, 2021.

Myocardial Perfusion Imaging After Severe COVID-19 Infection Demonstrates Regional Ischemia Rather Than Global Blood Flow Reduction.

Background: Acute myocardial damage is common in severe COVID-19. Post-mortem studies have implicated microvascular thrombosis, with cardiovascular magnetic resonance (CMR) demonstrating a high prevalence of myocardial infarction and myocarditis-like scar. The microcirculatory sequelae are incompletely characterized. Perfusion CMR can quantify the stress myocardial blood flow (MBF) and identify its association with infarction and myocarditis. Objectives: To determine the impact of the severe hospitalized COVID-19 on global and regional myocardial perfusion in recovered patients. Methods: A case-control study of previously hospitalized, troponin-positive COVID-19 patients was undertaken. The results were compared with a propensity-matched, pre-COVID chest pain cohort (referred for clinical CMR; angiography subsequently demonstrating unobstructed coronary arteries) and 27 healthy volunteers (HV). The analysis used visual assessment for the regional perfusion defects and AI-based segmentation to derive the global and regional stress and rest MBF. Results: Ninety recovered post-COVID patients {median age 64 [interquartile range (IQR) 54-71] years, 83% male, 44% requiring the intensive care unit (ICU)} underwent adenosine-stress perfusion CMR at a median of 61 (IQR 29-146) days post-discharge. The mean left ventricular ejection fraction (LVEF) was 67 ± 10%; 10 (11%) with impaired LVEF. Fifty patients (56%) had late gadolinium enhancement (LGE); 15 (17%) had infarct-pattern, 31 (34%) had non-ischemic, and 4 (4.4%) had mixed pattern LGE. Thirty-two patients (36%) had adenosine-induced regional perfusion defects, 26 out of 32 with at least one segment without prior infarction. The global stress MBF in post-COVID patients was similar to the age-, sex- and co-morbidities of the matched controls (2.53 ± 0.77 vs. 2.52 ± 0.79 ml/g/min, p = 0.10), though lower than HV (3.00 ± 0.76 ml/g/min, p< 0.01). Conclusions: After severe hospitalized COVID-19 infection, patients who attended clinical ischemia testing had little evidence of significant microvascular disease at 2 months post-discharge. The high prevalence of regional inducible ischemia and/or infarction (nearly 40%) may suggest that occult coronary disease is an important putative mechanism for troponin elevation in this cohort. This should be considered hypothesis-generating for future studies which combine ischemia and anatomical assessment.

The impact of artificial intelligence on clinical education: perceptions of postgraduate trainee doctors in London (UK) and recommendations for trainers.

BACKGROUND: Artificial intelligence (AI) technologies are increasingly used in clinical practice. Although there is robust evidence that AI innovations can improve patient care, reduce clinicians' workload and increase efficiency, their impact on medical training and education remains unclear. METHODS: A survey of trainee doctors' perceived impact of AI technologies on clinical training and education was conducted at UK NHS postgraduate centers in London between October and December 2020. Impact assessment mirrored domains in training curricula such as 'clinical judgement', 'practical skills' and 'research and quality improvement skills'. Significance between Likert-type data was analysed using Fisher's exact test. Response variations between clinical specialities were analysed using k-modes clustering. Free-text responses were analysed by thematic analysis. RESULTS: Two hundred ten doctors responded to the survey (response rate 72%). The majority (58%) perceived an overall positive impact of AI technologies on their training and education. Respondents agreed that AI would reduce clinical workload (62%) and improve research and audit training (68%). Trainees were skeptical that it would improve clinical judgement (46% agree, p = 0.12) and practical skills training (32% agree, p < 0.01). The majority reported insufficient AI training in their current curricula (92%), and supported having more formal AI training (81%). CONCLUSIONS: Trainee doctors have an overall positive perception of AI technologies' impact on clinical training. There is optimism that it will improve 'research and quality improvement' skills and facilitate 'curriculum mapping'. There is skepticism that it may reduce educational opportunities to develop 'clinical judgement' and 'practical skills'. Medical educators should be mindful that these domains are protected as AI develops. We recommend that 'Applied AI' topics are formalized in curricula and digital technologies leveraged to deliver clinical education.

Neural architecture search of echocardiography view classifiers.

Purpose: Echocardiography is the most commonly used modality for assessing the heart in clinical practice. In an echocardiographic exam, an ultrasound probe samples the heart from different orientations and positions, thereby creating different viewpoints for assessing the cardiac function. The determination of the probe viewpoint forms an essential step in automatic echocardiographic image analysis. Approach: In this study, convolutional neural networks are used for the automated identification of 14 different anatomical echocardiographic views (larger than any previous study) in a dataset of 8732 videos acquired from 374 patients. Differentiable architecture search approach was utilized to design small neural network architectures for rapid inference while maintaining high accuracy. The impact of the image quality and resolution, size of the training dataset, and number of echocardiographic view classes on the efficacy of the models were also investigated. Results: In contrast to the deeper classification architectures, the proposed models had significantly lower number of trainable parameters (up to 99.9% reduction), achieved comparable classification performance (accuracy 88.4% to 96%, precision 87.8% to 95.2%, recall 87.1% to 95.1%) and real-time performance with inference time per image of 3.6 to 12.6 ms. Conclusion: Compared with the standard classification neural network architectures, the proposed models are faster and achieve comparable classification performance. They also require less training data. Such models can be used for real-time detection of the standard views.

Automated Left Ventricular Dimension Assessment Using Artificial Intelligence Developed and Validated by a UK-Wide Collaborative.

BACKGROUND: requires training and validation to standards expected of humans. We developed an online platform and established the Unity Collaborative to build a dataset of expertise from 17 hospitals for training, validation, and standardization of such techniques. METHODS: The training dataset consisted of 2056 individual frames drawn at random from 1265 parasternal long-axis video-loops of patients undergoing clinical echocardiography in 2015 to 2016. Nine experts labeled these images using our online platform. From this, we trained a convolutional neural network to identify keypoints. Subsequently, 13 experts labeled a validation dataset of the end-systolic and end-diastolic frame from 100 new video-loops, twice each. The 26-opinion consensus was used as the reference standard. The primary outcome was precision SD, the SD of the differences between AI measurement and expert consensus. RESULTS: In the validation dataset, the AI's precision SD for left ventricular internal dimension was 3.5 mm. For context, precision SD of individual expert measurements against the expert consensus was 4.4 mm. Intraclass correlation coefficient between AI and expert consensus was 0.926 (95% CI, 0.904-0.944), compared with 0.817 (0.778-0.954) between individual experts and expert consensus. For interventricular septum thickness, precision SD was 1.8 mm for AI (intraclass correlation coefficient, 0.809; 0.729-0.967), versus 2.0 mm for individuals (intraclass correlation coefficient, 0.641; 0.568-0.716). For posterior wall thickness, precision SD was 1.4 mm for AI (intraclass correlation coefficient, 0.535 [95% CI, 0.379-0.661]), versus 2.2 mm for individuals (0.366 [0.288-0.462]). We present all images and annotations. This highlights challenging cases, including poor image quality and tapered ventricles. CONCLUSIONS: Experts at multiple institutions successfully cooperated to build a collaborative AI. This performed as well as individual experts. Future echocardiographic AI research should use a consensus of experts as a reference. Our collaborative welcomes new partners who share our commitment to publish all methods, code, annotations, and results openly.

Aerosolised fluorescein can quantify FFP mask faceseal leakage: a cost-effective adaptation to the existing point of care fit test.

BACKGROUND: A qualitative fit test using bitter-tasting aerosols is the commonest way to determine filtering face-piece (FFP) mask leakage. This taste test is subjective and biased by placebo. We propose a cheap, quantitative modification of the taste test by measuring the amount of fluorescein stained filter paper behind the mask using image analysis. METHODS: A bitter-tasting fluorescein solution was aerosolised during mask fit tests, with filter paper placed on masks' inner surfaces. Participants reported whether they could taste bitterness to determine taste test 'pass' or 'fail' results. Filter paper photographs were digitally analysed to quantify total fluorescence (TF). RESULTS: Fifty-six healthcare professionals were fit tested; 32 (57%) 'passed' the taste test. TF between the taste test 'pass' and 'fail' groups was significantly different (p<0.001). A cut-off (TF = 5.0 × 106 units) was determined at precision (78%) and recall (84%), resulting in 5/56 participants (9%) reclassified from 'pass' to 'fail' by the fluorescein test. Seven out of 56 (12%) reclassified from 'fail' to 'pass'. CONCLUSION: Fluorescein is detectable and sensitive at identifying FFP mask leaks. These low-cost adaptations can enhance exiting fit testing to determine 'pass' and 'fail' groups, protecting those who 'passed' the taste test but have high fluorescein leak, and reassuring those who 'failed' the taste test despite having little fluorescein leak.