A three-dimensional left atrial motion estimation from retrospective gated computed tomography: application in heart failure patients with atrial fibrillation.

Background: A reduced left atrial (LA) strain correlates with the presence of atrial fibrillation (AF). Conventional atrial strain analysis uses two-dimensional (2D) imaging, which is, however, limited by atrial foreshortening and an underestimation of through-plane motion. Retrospective gated computed tomography (RGCT) produces high-fidelity three-dimensional (3D) images of the cardiac anatomy throughout the cardiac cycle that can be used for estimating 3D mechanics. Its feasibility for LA strain measurement, however, is understudied. Aim: The aim of this study is to develop and apply a novel workflow to estimate 3D LA motion and calculate the strain from RGCT imaging. The utility of global and regional strains to separate heart failure in patients with reduced ejection fraction (HFrEF) with and without AF is investigated. Methods: A cohort of 30 HFrEF patients with (n = 9) and without (n = 21) AF underwent RGCT prior to cardiac resynchronisation therapy. The temporal sparse free form deformation image registration method was optimised for LA feature tracking in RGCT images and used to estimate 3D LA endocardial motion. The area and fibre reservoir strains were calculated over the LA body. Universal atrial coordinates and a human atrial fibre atlas enabled the regional strain calculation and the fibre strain calculation along the local myofibre orientation, respectively. Results: It was found that global reservoir strains were significantly reduced in the HFrEF + AF group patients compared with the HFrEF-only group patients (area strain: 11.2 ± 4.8% vs. 25.3 ± 12.6%, P = 0.001; fibre strain: 4.5 ± 2.0% vs. 15.2 ± 8.8%, P = 0.001), with HFrEF + AF patients having a greater regional reservoir strain dyssynchrony. All regional reservoir strains were reduced in the HFrEF + AF patient group, in whom the inferior wall strains exhibited the most significant differences. The global reservoir fibre strain and LA volume + posterior wall reservoir fibre strain exceeded LA volume alone and 2D global longitudinal strain (GLS) for AF classification (area-under-the-curve: global reservoir fibre strain: 0.94 ± 0.02, LA volume + posterior wall reservoir fibre strain: 0.95 ± 0.02, LA volume: 0.89 ± 0.03, 2D GLS: 0.90 ± 0.03). Conclusion: RGCT enables 3D LA motion estimation and strain calculation that outperforms 2D strain metrics and LA enlargement for AF classification. Differences in regional LA strain could reflect regional myocardial properties such as atrial fibrosis burden.

Hybrid artificial intelligence outcome prediction using features extraction from stress perfusion cardiac magnetic resonance images and electronic health records.

Background: Prediction of clinical outcomes in coronary artery disease (CAD) has been conventionally achieved using clinical risk factors. The relationship between imaging features and outcome is still not well understood. This study aims to use artificial intelligence to link image features with mortality outcome. Methods: A retrospective study was performed on patients who had stress perfusion cardiac magnetic resonance (SP-CMR) between 2011 and 2021. The endpoint was all-cause mortality. Convolutional neural network (CNN) was used to extract features from stress perfusion images, and multilayer perceptron (MLP) to extract features from electronic health records (EHRs), both networks were concatenated in a hybrid neural network (HNN) to predict study endpoint. Image CNN was trained to predict study endpoint directly from images. HNN and image CNN were compared with a linear clinical model using area under the curve (AUC), F1 scores, and McNemar's test. Results: Total of 1,286 cases were identified, with 201 death events (16%). The clinical model had good performance (AUC =80%, F1 score =37%). Best Image CNN model showed AUC =72% and F1 score =38%. HNN outperformed the other two models (AUC =82%, F1 score =43%). McNemar's test showed statistical difference between image CNN and both clinical model (P<0.01) and HNN (P<0.01). There was no significant difference between HNN and clinical model (P=0.15). Conclusions: Death in patients with suspected or known CAD can be predicted directly from stress perfusion images without clinical knowledge. Prediction can be improved by HNN that combines clinical and SP-CMR images.

Adjuvant chemotherapy for adenocarcinoma arising from intraductal papillary mucinous neoplasia: multicentre ADENO-IPMN study.

BACKGROUND: The clinical impact of adjuvant chemotherapy after resection for adenocarcinoma arising from intraductal papillary mucinous neoplasia is unclear. The aim of this study was to identify factors related to receipt of adjuvant chemotherapy and its impact on recurrence and survival. METHODS: This was a multicentre retrospective study of patients undergoing pancreatic resection for adenocarcinoma arising from intraductal papillary mucinous neoplasia between January 2010 and December 2020 at 18 centres. Recurrence and survival outcomes for patients who did and did not receive adjuvant chemotherapy were compared using propensity score matching. RESULTS: Of 459 patients who underwent pancreatic resection, 275 (59.9%) received adjuvant chemotherapy (gemcitabine 51.3%, gemcitabine-capecitabine 21.8%, FOLFIRINOX 8.0%, other 18.9%). Median follow-up was 78 months. The overall recurrence rate was 45.5% and the median time to recurrence was 33 months. In univariable analysis in the matched cohort, adjuvant chemotherapy was not associated with reduced overall (P = 0.713), locoregional (P = 0.283) or systemic (P = 0.592) recurrence, disease-free survival (P = 0.284) or overall survival (P = 0.455). Adjuvant chemotherapy was not associated with reduced site-specific recurrence. In multivariable analysis, there was no association between adjuvant chemotherapy and overall recurrence (HR 0.89, 95% c.i. 0.57 to 1.40), disease-free survival (HR 0.86, 0.59 to 1.30) or overall survival (HR 0.77, 0.50 to 1.20). Adjuvant chemotherapy was not associated with reduced recurrence in any high-risk subgroup (for example, lymph node-positive, higher AJCC stage, poor differentiation). No particular chemotherapy regimen resulted in superior outcomes. CONCLUSION: Chemotherapy following resection of adenocarcinoma arising from intraductal papillary mucinous neoplasia does not appear to influence recurrence rates, recurrence patterns or survival.

Long-term Outcomes following Resection of Adenocarcinoma Arising from Intraductal Papillary Mucinous Neoplasm (A-IPMN) versus Pancreatic Ductal Adenocarcinoma (PDAC): A Propensity-score Matched Analysis.

OBJECTIVE: The aim of the present study was to compare long-term post-resection oncological outcomes between A-IPMN and PDAC. SUMMARY BACKGROUND DATA: Knowledge of long term oncological outcomes (e.g recurrence and survival data) comparing between adenocarcinoma arising from intraductal papillary mucinous neoplasms (A-IPMN) and pancreatic ductal adenocarcinoma (PDAC) is scarce. METHODS: Patients undergoing pancreatic resection (2010-2020) for A-IPMN were identified retrospectively from 18 academic pancreatic centres and compared with PDAC patients from the same time-period. Propensity-score matching (PSM) was performed and survival and recurrence were compared between A-IPMN and PDAC. RESULTS: 459 A-IPMN patients (median age,70; M:F,250:209) were compared with 476 PDAC patients (median age,69; M:F,262:214). A-IPMN patients had lower T-stage, lymphovascular invasion (51.4%vs. 75.6%), perineural invasion (55.8%vs. 71.2%), lymph node positivity (47.3vs. 72.3%) and R1 resection (38.6%vs. 56.3%) compared to PDAC(P<0.001). The median survival and time-to-recurrence for A-IPMN versus PDAC were 39.0 versus19.5months (P<0.001) and 33.1 versus 14.8months (P<0.001), respectively (median follow-up,78 vs.73 months). Ten-year overall survival for A-IPMN was 34.6%(27/78) and PDAC was 9%(6/67). A-IPMN had higher rates of peritoneal (23.0 vs. 9.1%, P<0.001) and lung recurrence (27.8% vs. 15.6%, P<0.001) but lower rates of locoregional recurrence (39.7% vs. 57.8%; P<0.001). Matched analysis demonstrated inferior overall survival (P=0.005), inferior disease-free survival (P=0.003) and higher locoregional recurrence (P<0.001) in PDAC compared to A-IPMN but no significant difference in systemic recurrence rates (P=0.695). CONCLUSIONS: PDACs have inferior survival and higher recurrence rates compared to A-IPMN in matched cohorts. Locoregional recurrence is higher in PDAC but systemic recurrence rates are comparable and constituted by their own distinctive site-specific recurrence patterns.

A universal biventricular coordinate system incorporating valve annuli: Validation in congenital heart disease.

Universal coordinate systems have been proposed to facilitate anatomic registration between three-dimensional images, data and models of the ventricles of the heart. However, current universal ventricular coordinate systems do not account for the outflow tracts and valve annuli where the anatomy is complex. Here we propose an extension to the 'Cobiveco' biventricular coordinate system that also accounts for the intervalvular bridges of the base and provides a tool for anatomically consistent registration between widely varying biventricular shapes. CobivecoX uses a novel algorithm to separate intervalvular bridges and assign new coordinates, including an inflow-outflow coordinate, to describe local positions in these regions uniquely and consistently. Anatomic consistency of registration was validated using curated three-dimensional biventricular shape models derived from cardiac MRI measurements in normal hearts and hearts from patients with congenital heart diseases. This new method allows the advantages of universal cardiac coordinates to be used for three-dimensional ventricular imaging data and models that include the left and right ventricular outflow tracts and valve annuli.

A Deep Learning-Based Integrated Framework for Quality-Aware Undersampled Cine Cardiac MRI Reconstruction and Analysis.

Cine cardiac magnetic resonance (CMR) imaging is considered the gold standard for cardiac function evaluation. However, cine CMR acquisition is inherently slow and in recent decades considerable effort has been put into accelerating scan times without compromising image quality or the accuracy of derived results. In this article, we present a fully-automated, quality-controlled integrated framework for reconstruction, segmentation and downstream analysis of undersampled cine CMR data. The framework produces high quality reconstructions and segmentations, leading to undersampling factors that are optimised on a scan-by-scan basis. This results in reduced scan times and automated analysis, enabling robust and accurate estimation of functional biomarkers. To demonstrate the feasibility of the proposed approach, we perform simulations of radial k-space acquisitions using in-vivo cine CMR data from 270 subjects from the UK Biobank (with synthetic phase) and in-vivo cine CMR data from 16 healthy subjects (with real phase). The results demonstrate that the optimal undersampling factor varies for different subjects by approximately 1 to 2 seconds per slice. We show that our method can produce quality-controlled images in a mean scan time reduced from 12 to 4 seconds per slice, and that image quality is sufficient to allow clinically relevant parameters to be automatically estimated to lie within 5% mean absolute difference.

Developing Cardiac Digital Twins at Scale: Insights from Personalised Myocardial Conduction Velocity.

Large-cohort studies using cardiovascular imaging and diagnostic datasets have assessed cardiac anatomy, function, and outcomes, but typically do not reveal underlying biological mechanisms. Cardiac digital twins (CDTs) provide personalized physics- and physiology-constrained in-silico representations, enabling inference of multi-scale properties tied to these mechanisms. We constructed 3464 anatomically-accurate CDTs using cardiac magnetic resonance images from UK biobank and personalised their myocardial conduction velocities (CVs) from electrocardiograms (ECG), through an automated framework. We found well-known sex-specific differences in QRS duration were fully explained by myocardial anatomy, as CV remained consistent across sexes. Conversely, significant associations of CV with ageing and increased BMI suggest myocardial tissue remodelling. Novel associations were observed with left ventricular ejection fraction and mental-health phenotypes, through a phenome-wide association study, and CV was also linked with adverse clinical outcomes. Our study highlights the utility of population-based CDTs in assessing intersubject variability and uncovering strong links with mental health.

lGeneralized super-resolution 4D Flow MRI-using ensemble learning to extend across the cardiovascular system.

4D Flow Magnetic Resonance Imaging (4D Flow MRI) is a non-invasive measurement technique capable of quantifying blood flow across the cardiovascular system. While practical use is limited by spatial resolution and image noise, incorporation of trained super-resolution (SR) networks has potential to enhance image quality post-scan. However, these efforts have predominantly been restricted to narrowly defined cardiovascular domains, with limited exploration of how SR performance extends across the cardiovascular system; a task aggravated by contrasting hemodynamic conditions apparent across the cardiovasculature. The aim of our study was to explore the generalizability of SR 4D Flow MRI using a combination of heterogeneous training sets and dedicated ensemble learning. With synthetic training data generated across three disparate domains (cardiac, aortic, cerebrovascular), varying convolutional base and ensemble learners were evaluated as a function of domain and architecture, quantifying performance on both in-silico and acquired in-vivo data from the same three domains. Results show that both bagging and stacking ensembling enhance SR performance across domains, accurately predicting high-resolution velocities from low-resolution input data in-silico. Likewise, optimized networks successfully recover native resolution velocities from downsampled in-vivo data, as well as show qualitative potential in generating denoised SR-images from clinicallevel input data. In conclusion, our work presents a viable approach for generalized SR 4D Flow MRI, with ensemble learning extending utility across various clinical areas of interest.

Combining generative modelling and semi-supervised domain adaptation for whole heart cardiovascular magnetic resonance angiography segmentation.

BACKGROUND: Quantification of three-dimensional (3D) cardiac anatomy is important for the evaluation of cardiovascular diseases. Changes in anatomy are indicative of remodeling processes as the heart tissue adapts to disease. Although robust segmentation methods exist for computed tomography angiography (CTA), few methods exist for whole-heart cardiovascular magnetic resonance angiograms (CMRA) which are more challenging due to variable contrast, lower signal to noise ratio and a limited amount of labeled data. METHODS: Two state-of-the-art unsupervised generative deep learning domain adaptation architectures, generative adversarial networks and variational auto-encoders, were applied to 3D whole heart segmentation of both conventional (n = 20) and high-resolution (n = 45) CMRA (target) images, given segmented CTA (source) images for training. An additional supervised loss function was implemented to improve performance given 10%, 20% and 30% segmented CMRA cases. A fully supervised nn-UNet trained on the given CMRA segmentations was used as the benchmark. RESULTS: The addition of a small number of segmented CMRA training cases substantially improved performance in both generative architectures in both standard and high-resolution datasets. Compared with the nn-UNet benchmark, the generative methods showed substantially better performance in the case of limited labelled cases. On the standard CMRA dataset, an average 12% (adversarial method) and 10% (variational method) improvement in Dice score was obtained. CONCLUSIONS: Unsupervised domain-adaptation methods for CMRA segmentation can be boosted by the addition of a small number of supervised target training cases. When only few labelled cases are available, semi-supervised generative modelling is superior to supervised methods.

Risk of Recurrence after Surgical Resection for Adenocarcinoma Arising from Intraductal Papillary Mucinous Neoplasia (IPMN) with Patterns of Distribution and Treatment: An International, Multicentre, Observational Study.

OBJECTIVE: This international multicentre cohort study aims to identify recurrence patterns and treatment of first and second recurrence in a large cohort of patients after pancreatic resection for adenocarcinoma arising from IPMN. SUMMARY BACKGROUND DATA: Recurrence patterns and treatment of recurrence post resection of adenocarcinoma arising from IPMN are poorly explored. METHOD: Patients undergoing pancreatic resection for adenocarcinoma from IPMN between January 2010 to December 2020 at 18 pancreatic centres were identified. Survival analysis was performed by the Kaplan-Meier log rank test and multivariable logistic regression by Cox-Proportional Hazards modelling. Endpoints were recurrence (time-to, location, and pattern of recurrence) and survival (overall survival and adjusted for treatment provided). RESULTS: Four hundred and fifty-nine patients were included (median, 70 y; IQR, 64-76; male, 54 percent) with a median follow-up of 26.3 months (IQR, 13.0-48.1 mo). Recurrence occurred in 209 patients (45.5 percent; median time to recurrence, 32.8 months, early recurrence [within 1 y], 23.2 percent). Eighty-three (18.1 percent) patients experienced a local regional recurrence and 164 (35.7 percent) patients experienced distant recurrence. Adjuvant chemotherapy was not associated with reduction in recurrence (HR 1.09;P=0.669) One hundred and twenty patients with recurrence received further treatment. The median survival with and without additional treatment was 27.0 and 14.6 months (P<0.001), with no significant difference between treatment modalities. There was no significant difference in survival between location of recurrence (P=0.401). CONCLUSION: Recurrence after pancreatic resection for adenocarcinoma arising from IPMN is frequent with a quarter of patients recurring within 12 months. Treatment of recurrence is associated with improved overall survival and should be considered.

Quality of life after severe acute pancreatitis: systematic review.

BACKGROUND: Severe acute pancreatitis, the most severe form of acute pancreatitis, can alter pancreatic morphology, physiology, and function resulting in long-term morbidity, even after a single episode. This review assesses long-term outcomes and quality of life of severe acute pancreatitis. METHODS: A comprehensive literature review was conducted across MEDLINE, Embase, Scopus, and PubMed electronic databases on 18 January 2021 and updated on 26 April 2022 to ensure no new literature had been omitted. All studies were prospective or retrospective, included adult patients (>18 years) presenting with acute pancreatitis for whom data on long-term outcomes specifically after severe acute pancreatitis were reported. Quantitative and qualitative data extraction and synthesis were carried out and no meta-analysis was performed. Outcome measures included aetiology and mortality of severe acute pancreatitis, length of stay, endocrine and exocrine pancreatic insufficiency, chronic symptoms, and quality of life compared with healthy controls as assessed by validated questionnaires. RESULTS: Fourteen retrospective cohort studies were included, for a total of 779 patients, using quality of life questionnaires. The most common aetiology of severe acute pancreatitis was biliary (36 per cent) followed by alcoholic (29 per cent). Mortality rate ranged from 5 to 35 per cent and length of stay ranged from 2 to 367 days. Quality of life was somewhat lower in patients with exocrine insufficiency, but unaffected by endocrine insufficiency or chronic symptoms. Quality of life was more likely to be reduced in the first 4 years but normalize thereafter and was more likely to be negatively affected where alcohol was the aetiology. In four studies, the relationship between disease severity and lower quality of life was investigated, and a significant correlation was found. CONCLUSION: The review shows how a single episode of severe acute pancreatitis can have a variable effect on long-term quality of life, which is different to previous studies showing a strong reduction in quality of life. This could indicate that in current times treatment modalities are more effective.

Association of Longer Leukocyte Telomere Length With Cardiac Size, Function, and Heart Failure.

Importance: Longer leukocyte telomere length (LTL) is associated with a lower risk of adverse cardiovascular outcomes. The extent to which variation in LTL is associated with intermediary cardiovascular phenotypes is unclear. Objective: To evaluate the associations between LTL and a diverse set of cardiovascular imaging phenotypes. Design, Setting, and Participants: This is a population-based cross-sectional study of UK Biobank participants recruited from 2006 to 2010. LTL was measured using a quantitative polymerase chain reaction method. Cardiovascular measurements were derived from cardiovascular magnetic resonance using machine learning. The median (IQR) duration of follow-up was 12.0 (11.3-12.7) years. The associations of LTL with imaging measurements and incident heart failure (HF) were evaluated by multivariable regression models. Genetic associations between LTL and significantly associated traits were investigated by mendelian randomization. Data were analyzed from January to May 2023. Exposure: LTL. Main Outcomes and Measures: Cardiovascular imaging traits and HF. Results: Of 40 459 included participants, 19 529 (48.3%) were men, and the mean (SD) age was 55.1 (7.6) years. Longer LTL was independently associated with a pattern of positive cardiac remodeling (higher left ventricular mass, larger global ventricular size and volume, and higher ventricular and atrial stroke volumes) and a lower risk of incident HF (LTL fourth quartile vs first quartile: hazard ratio, 0.86; 95% CI, 0.81-0.91; P = 1.8 × 10-6). Mendelian randomization analysis suggested a potential causal association between LTL and left ventricular mass, global ventricular volume, and left ventricular stroke volume. Conclusions and Relevance: In this cross-sectional study, longer LTL was associated with a larger heart with better cardiac function in middle age, which could potentially explain the observed lower risk of incident HF.

Quinapril treatment curtails decline of global longitudinal strain and mechanical function in hypertensive rats.

BACKGROUND: Left ventricular (LV) global longitudinal strain (GLS) has been proposed as an early imaging biomarker of cardiac mechanical dysfunction. OBJECTIVE: To assess the impact of angiotensin-converting enzyme (ACE) inhibitor treatment of hypertensive heart disease on LV GLS and mechanical function. METHODS: The spontaneously hypertensive rat (SHR) model of hypertensive heart disease ( n  = 38) was studied. A subset of SHRs received quinapril (TSHR, n  = 16) from 3 months (mo). Wistar Kyoto rats (WKY, n  = 13) were used as controls. Tagged cardiac MRI was performed using a 4.7 T Varian preclinical scanner. RESULTS: The SHRs had significantly lower LV ejection fraction (EF) than the WKYs at 3 mo (53.0 ±â€Š1.7% vs. 69.6 ±â€Š2.1%, P  < 0.05), 14 mo (57.0 ±â€Š2.5% vs. 74.4 ±â€Š2.9%, P  < 0.05) and 24 mo (50.1 ±â€Š2.4% vs. 67.0 ±â€Š2.0%, P  < 0.01). At 24 mo, ACE inhibitor treatment was associated with significantly greater LV EF in TSHRs compared to untreated SHRs (64.2 ±â€Š3.4% vs. 50.1 ±â€Š2.4%, P  < 0.01). Peak GLS magnitude was significantly lower in SHRs compared with WKYs at 14 months (7.5% ±â€Š0.4% vs. 9.9 ±â€Š0.8%, P  < 0.05). At 24 months, Peak GLS magnitude was significantly lower in SHRs compared with both WKYs (6.5 ±â€Š0.4% vs. 9.7 ±â€Š1.0%, P  < 0.01) and TSHRs (6.5 ±â€Š0.4% vs. 9.6 ±â€Š0.6%, P  < 0.05). CONCLUSIONS: ACE inhibitor treatment curtails the decline in global longitudinal strain in hypertensive rats, with the treatment group exhibiting significantly greater LV EF and GLS magnitude at 24 mo compared with untreated SHRs.

Correcting bias in cardiac geometries derived from multimodal images using spatiotemporal mapping.

Cardiovascular imaging studies provide a multitude of structural and functional data to better understand disease mechanisms. While pooling data across studies enables more powerful and broader applications, performing quantitative comparisons across datasets with varying acquisition or analysis methods is problematic due to inherent measurement biases specific to each protocol. We show how dynamic time warping and partial least squares regression can be applied to effectively map between left ventricular geometries derived from different imaging modalities and analysis protocols to account for such differences. To demonstrate this method, paired real-time 3D echocardiography (3DE) and cardiac magnetic resonance (CMR) sequences from 138 subjects were used to construct a mapping function between the two modalities to correct for biases in left ventricular clinical cardiac indices, as well as regional shape. Leave-one-out cross-validation revealed a significant reduction in mean bias, narrower limits of agreement, and higher intraclass correlation coefficients for all functional indices between CMR and 3DE geometries after spatiotemporal mapping. Meanwhile, average root mean squared errors between surface coordinates of 3DE and CMR geometries across the cardiac cycle decreased from 7 ± 1 to 4 ± 1 mm for the total study population. Our generalised method for mapping between time-varying cardiac geometries obtained using different acquisition and analysis protocols enables the pooling of data between modalities and the potential for smaller studies to leverage large population databases for quantitative comparisons.

Automated segmentation of long and short axis DENSE cardiovascular magnetic resonance for myocardial strain analysis using spatio-temporal convolutional neural networks.

BACKGROUND: Cine Displacement Encoding with Stimulated Echoes (DENSE) facilitates the quantification of myocardial deformation, by encoding tissue displacements in the cardiovascular magnetic resonance (CMR) image phase, from which myocardial strain can be estimated with high accuracy and reproducibility. Current methods for analyzing DENSE images still heavily rely on user input, making this process time-consuming and subject to inter-observer variability. The present study sought to develop a spatio-temporal deep learning model for segmentation of the left-ventricular (LV) myocardium, as spatial networks often fail due to contrast-related properties of DENSE images. METHODS: 2D + time nnU-Net-based models have been trained to segment the LV myocardium from DENSE magnitude data in short- and long-axis images. A dataset of 360 short-axis and 124 long-axis slices was used to train the networks, from a combination of healthy subjects and patients with various conditions (hypertrophic and dilated cardiomyopathy, myocardial infarction, myocarditis). Segmentation performance was evaluated using ground-truth manual labels, and a strain analysis using conventional methods was performed to assess strain agreement with manual segmentation. Additional validation was performed using an externally acquired dataset to compare the inter- and intra-scanner reproducibility with respect to conventional methods. RESULTS: Spatio-temporal models gave consistent segmentation performance throughout the cine sequence, while 2D architectures often failed to segment end-diastolic frames due to the limited blood-to-myocardium contrast. Our models achieved a DICE score of 0.83 ± 0.05 and a Hausdorff distance of 4.0 ± 1.1 mm for short-axis segmentation, and 0.82 ± 0.03 and 7.9 ± 3.9 mm respectively for long-axis segmentations. Strain measurements obtained from automatically estimated myocardial contours showed good to excellent agreement with manual pipelines, and remained within the limits of inter-user variability estimated in previous studies. CONCLUSION: Spatio-temporal deep learning shows increased robustness for the segmentation of cine DENSE images. It provides excellent agreement with manual segmentation for strain extraction. Deep learning will facilitate the analysis of DENSE data, bringing it one step closer to clinical routine.

A deep learning approach for fully automated cardiac shape modeling in tetralogy of Fallot.

BACKGROUND: Cardiac shape modeling is a useful computational tool that has provided quantitative insights into the mechanisms underlying dysfunction in heart disease. The manual input and time required to make cardiac shape models, however, limits their clinical utility. Here we present an end-to-end pipeline that uses deep learning for automated view classification, slice selection, phase selection, anatomical landmark localization, and myocardial image segmentation for the automated generation of three-dimensional, biventricular shape models. With this approach, we aim to make cardiac shape modeling a more robust and broadly applicable tool that has processing times consistent with clinical workflows. METHODS: Cardiovascular magnetic resonance (CMR) images from a cohort of 123 patients with repaired tetralogy of Fallot (rTOF) from two internal sites were used to train and validate each step in the automated pipeline. The complete automated pipeline was tested using CMR images from a cohort of 12 rTOF patients from an internal site and 18 rTOF patients from an external site. Manually and automatically generated shape models from the test set were compared using Euclidean projection distances, global ventricular measurements, and atlas-based shape mode scores. RESULTS: The mean absolute error (MAE) between manually and automatically generated shape models in the test set was similar to the voxel resolution of the original CMR images for end-diastolic models (MAE = 1.9 ± 0.5 mm) and end-systolic models (MAE = 2.1 ± 0.7 mm). Global ventricular measurements computed from automated models were in good agreement with those computed from manual models. The average mean absolute difference in shape mode Z-score between manually and automatically generated models was 0.5 standard deviations for the first 20 modes of a reference statistical shape atlas. CONCLUSIONS: Using deep learning, accurate three-dimensional, biventricular shape models can be reliably created. This fully automated end-to-end approach dramatically reduces the manual input required to create shape models, thereby enabling the rapid analysis of large-scale datasets and the potential to deploy statistical atlas-based analyses in point-of-care clinical settings. Training data and networks are available from cardiacatlas.org.

AI-AIF: artificial intelligence-based arterial input function for quantitative stress perfusion cardiac magnetic resonance.

Aims: One of the major challenges in the quantification of myocardial blood flow (MBF) from stress perfusion cardiac magnetic resonance (CMR) is the estimation of the arterial input function (AIF). This is due to the non-linear relationship between the concentration of gadolinium and the MR signal, which leads to signal saturation. In this work, we show that a deep learning model can be trained to predict the unsaturated AIF from standard images, using the reference dual-sequence acquisition AIFs (DS-AIFs) for training. Methods and results: A 1D U-Net was trained, to take the saturated AIF from the standard images as input and predict the unsaturated AIF, using the data from 201 patients from centre 1 and a test set comprised of both an independent cohort of consecutive patients from centre 1 and an external cohort of patients from centre 2 (n = 44). Fully-automated MBF was compared between the DS-AIF and AI-AIF methods using the Mann-Whitney U test and Bland-Altman analysis. There was no statistical difference between the MBF quantified with the DS-AIF [2.77 mL/min/g (1.08)] and predicted with the AI-AIF (2.79 mL/min/g (1.08), P = 0.33. Bland-Altman analysis shows minimal bias between the DS-AIF and AI-AIF methods for quantitative MBF (bias of -0.11 mL/min/g). Additionally, the MBF diagnosis classification of the AI-AIF matched the DS-AIF in 669/704 (95%) of myocardial segments. Conclusion: Quantification of stress perfusion CMR is feasible with a single-sequence acquisition and a single contrast injection using an AI-based correction of the AIF.