Dataset Information

Deep learning to diagnose cardiac amyloidosis from cardiovascular magnetic resonance.

ABSTRACT: Background: Cardiovascular magnetic resonance (CMR) is part of the diagnostic work-up for cardiac amyloidosis (CA). Deep learning (DL) is an application of artificial intelligence that may allow to automatically analyze CMR findings and establish the likelihood of CA.

Methods: 1.5 T CMR was performed in 206 subjects with suspected CA (n?=?100, 49% with unexplained left ventricular (LV) hypertrophy; n?=?106, 51% with blood dyscrasia and suspected light-chain amyloidosis). Patients were randomly assigned to the training (n?=?134, 65%), validation (n?=?30, 15%), and testing subgroups (n?=?42, 20%). Short axis, 2-chamber, 4-chamber late gadolinium enhancement (LGE) images were evaluated by 3 networks (DL algorithms). The tags "amyloidosis present" or "absent" were attributed when the average probability of CA from the 3 networks was???50% or?
Results: The DL strategy displayed good diagnostic accuracy (88%), with an area under the curve (AUC) of 0.982. The precision (positive predictive value), recall score (sensitivity), and F1 score (a measure of test accuracy) were 83%, 95%, and 89% respectively. A ML algorithm considering all CMR features had a similar diagnostic yield to DL strategy (AUC 0.952 vs. 0.982; p?=?0.39).

Conclusions: A DL approach evaluating LGE acquisitions displayed a similar diagnostic performance for CA to a ML-based approach, which simulates CMR reading by experienced operators.

SUBMITTER: Martini N

PROVIDER: S-EPMC7720569 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep learning to diagnose cardiac amyloidosis from cardiovascular magnetic resonance.

Martini Nicola N Aimo Alberto A Barison Andrea A Della Latta Daniele D Vergaro Giuseppe G Aquaro Giovanni Donato GD Ripoli Andrea A Emdin Michele M Chiappino Dante D

Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance 20201207 1

<h4>Background</h4>Cardiovascular magnetic resonance (CMR) is part of the diagnostic work-up for cardiac amyloidosis (CA). Deep learning (DL) is an application of artificial intelligence that may allow to automatically analyze CMR findings and establish the likelihood of CA.<h4>Methods</h4>1.5 T CMR was performed in 206 subjects with suspected CA (n = 100, 49% with unexplained left ventricular (LV) hypertrophy; n = 106, 51% with blood dyscrasia and suspected light-chain amyloidosis). Patients we ...[more]

PMID: 33287829

Similar Datasets

Project description:Background Cardiac amyloidosis (CA) is a disease of interstitial myocardial infiltration, usually by light chains or transthyretin. We used diffusion tensor cardiovascular magnetic resonance (DT-CMR) to noninvasively assess the effects of amyloid infiltration on the cardiac microstructure. Methods DT-CMR was performed at diastole and systole in 20 CA, 11 hypertrophic cardiomyopathy, and 10 control subjects with calculation of mean diffusivity, fractional anisotropy, and sheetlet orientation (secondary eigenvector angle). Results Mean diffusivity was elevated and fractional anisotropy reduced in CA compared with both controls and hypertrophic cardiomyopathy (P<0.001). In CA, mean diffusivity was correlated with extracellular volume (r=0.68, P=0.004), and fractional anisotropy was inversely correlated with circumferential strain (r=-0.65, P=0.02). In CA, diastolic secondary eigenvector angle was elevated, and secondary eigenvector angle mobility was reduced compared with controls (both P<0.001). Diastolic secondary eigenvector angle was correlated with amyloid burden measured by extracellular volume in transthyretin, but not light chain amyloidosis. Conclusions DT-CMR can characterize the microstructural effects of amyloid infiltration and is a contrast-free method to identify the location and extent of the expanded disorganized myocardium. The diffusion biomarkers mean diffusivity and fractional anisotropy effectively discriminate CA from hypertrophic cardiomyopathy. DT-CMR demonstrated that failure of sheetlet relaxation in diastole correlated with extracellular volume in transthyretin, but not light chain amyloidosis. This indicates that different mechanisms may be responsible for impaired contractility in CA, with an amyloid burden effect in transthyretin, but an idiosyncratic effect in light chain amyloidosis. Consequently, DT-CMR offers a contrast-free tool to identify novel pathophysiology, improve diagnostics, and monitor disease through noninvasive microstructural assessment.

Project description:BackgroundIn-vivo cardiovascular magnetic resonance (CMR) diffusion tensor imaging (DTI) allows imaging of alterations of cardiac fiber architecture in diseased hearts. Cardiac amyloidosis (CA) causes myocardial infiltration of misfolded proteins with unknown consequences for myocardial microstructure. This study applied CMR DTI in CA to assess microstructural alterations and their consequences for myocardial function compared to healthy controls.MethodsTen patients with CA (8 AL, 2 ATTR) and ten healthy controls were studied using a diffusion-weighed second-order motion-compensated spin-echo sequence at 1.5 T. Additionally, left ventricular morphology, ejection fraction, strain and native T1 values were obtained in all subjects. In CA patients, T1 mapping was repeated after the administration of gadolinium for extracellular volume fraction (ECV) calculation. CMR DTI analysis was performed to yield the scalar diffusion metrics mean diffusivity (MD) and fractional anisotropy (FA) as well as the characteristics of myofiber orientation including helix, transverse and E2A sheet angle (HA, TA, E2A).ResultsMD and FA were found to be significantly different between CA patients and healthy controls (MD 1.77 ± 0.17 10- 3 vs 1.41 ± 0.07 10- 3 mm2/s, p < 0.001; FA 0.25 ± 0.04 vs 0.35 ± 0.03, p < 0.001). MD demonstrated an excellent correlation with native T1 (r = 0.908, p < 0.001) while FA showed a significant correlation with ECV in the CA population (r = - 0.851, p < 0.002). HA exhibited a more circumferential orientation of myofibers in CA patients, in conjunction with a higher TA standard deviation and a higher absolute E2A sheet angle. The transmural HA slope was found to be strongly correlated with the global longitudinal strain (r = 0.921, p < 0.001).ConclusionCMR DTI reveals significant alterations of scalar diffusion metrics in CA patients versus healthy controls. Elevated MD and lower FA values indicate myocardial disarray with higher diffusion in CA that correlates well with native T1 and ECV measures. In CA patients, CMR DTI showed pronounced circumferential orientation of the myofibers, which may provide the rationale for the reduction of global longitudinal strain that occurs in amyloidosis patients. Accordingly, CMR DTI captures specific features of amyloid infiltration, which provides a deeper understanding of the microstructural consequences of CA.

Project description:BackgroundCardiac amyloidosis (CA) is an infiltrative disease characterised by accumulation of amyloid deposits in the extracellular space of the myocardium-comprising transthyretin (ATTR) and light chain (AL) amyloidosis as the most frequent subtypes. Histopathological proof of amyloid deposits by endomyocardial biopsy (EMB) is the gold standard for diagnosis of CA. Cardiovascular magnetic resonance (CMR) allows non-invasive workup of suspected CA. We conducted a multi-centre study to assess the diagnostic value of CMR in comparison to EMB for the diagnosis of CA.MethodsWe studied N = 160 patients characterised by symptoms of heart failure and presence of left ventricular (LV) hypertrophy of unknown origin who presented to specialised cardiomyopathy centres in Germany and underwent further diagnostic workup by both CMR and EMB. If CA was diagnosed, additional subtyping based on EMB specimens and monoclonal protein studies in serum was performed. The CMR protocol comprised cine- and late-gadolinium-enhancement (LGE)-imaging as well as native and post-contrast T1-mapping (in a subgroup)-allowing to measure extracellular volume fraction (ECV) of the myocardium.ResultsAn EMB-based diagnosis of CA was made in N = 120 patients (CA group) whereas N = 40 patients demonstrated other diagnoses (CONTROL group). In the CA group, N = 114 (95%) patients showed a characteristic pattern of LGE indicative of CA. In the CONTROL group, only 1/40 (2%) patient showed a "false-positive" LGE pattern suggestive of CA. In the CA group, there was no patient with elevated T1-/ECV-values without a characteristic pattern of LGE indicative of CA. LGE-CMR showed a sensitivity of 95% and a specificity of 98% for the diagnosis of CA. The combination of a characteristic LGE pattern indicating CA with unremarkable monoclonal protein studies resulted in the diagnosis of ATTR-CA (confirmed by EMB) with a specificity of 98% [95%-confidence interval (CI) 92-100%] and a positive predictive value (PPV) of 99% (95%-CI 92-100%), respectively. The EMB-associated risk of complications was 3.13% in this study-without any detrimental or persistent complications.ConclusionNon-invasive CMR shows an excellent diagnostic accuracy and yield regarding CA. When combined with monoclonal protein studies, CMR can differentiate ATTR from AL with high accuracy and predictive value. However, invasive EMB remains a safe invasive gold-standard and allows to differentiate CA from other cardiomyopathies that can also cause LV hypertrophy.

Project description:OBJECTIVES:The purpose of this study was to explore the diagnostic usefulness of hybrid cardiac magnetic resonance (CMR) and positron emission tomography (PET) using 18F-fluorodeoxyglucose (FDG) for active cardiac sarcoidosis. BACKGROUND:Active cardiac sarcoidosis (aCS) is underdiagnosed and has a high mortality. METHODS:Patients with clinical suspicion of aCS underwent hybrid CMR/PET with late gadolinium enhancement (LGE) and FDG to assess the pattern of injury and disease activity, respectively. Patients were categorized visually as magnetic resonance (MR)+PET+ (characteristic LGE aligning exactly with increased FDG uptake), MR+PET- (characteristic LGE but no increased FDG), MR-PET- (neither characteristic LGE nor increased FDG), and MR-PET+ (increased FDG uptake in absence of characteristic LGE) and further characterized as aCS+ (MR+PET+) or aCS- (MR+PET-, MR-PET-, MR-PET+). FDG uptake was quantified using maximum target-to-normal-myocardium ratio and the net uptake rate (Ki) from dynamic Patlak analysis. Receiver-operating characteristic methods were used to identify imaging biomarkers for aCS. FDG PET was assessed using computed tomography/PET in 19 control subjects with healthy myocardium. RESULTS:A total of 25 patients (12 males; 54.9 ± 9.8 years of age) were recruited prospectively; 8 were MR+PET+, suggestive of aCS; 1 was MR+PET-, consistent with inactive cardiac sarcoidosis; and 8 were MR-PET-, with no imaging evidence of cardiac sarcoidosis. Eight patients were MR-PET+ (6 with global myocardial FDG uptake, 2 with focal-on-diffuse uptake); they demonstrated distinct Ki values and hyperintense maximum standardized uptake value compared with MR+PET+ patients. Similar hyperintense patterns of global (n = 9) and focal-on-diffuse (n = 2) FDG uptake were also observed in control patients, suggesting physiological myocardial uptake. Maximum target-to-normal-myocardium ratio values were higher in the aCS+ group (p < 0.001), demonstrating an area under the curve of 0.98 on receiver-operating characteristic analysis for the detection of aCS, with an optimal maximum target-to-normal myocardium ratio threshold of 1.2 (Youden index: 0.94). CONCLUSIONS:CMR/PET imaging holds major promise for the diagnosis of aCS, providing incremental information about both the pattern of injury and disease activity in a single scan. (In Vivo Molecular Imaging [MRI] of Atherothrombotic Lesions; NCT01418313).

Project description:The aim of this study is to shorten the breathhold and diastolic acquisition window in cardiac magnetic resonance fingerprinting (MRF) for simultaneous T1, T2, and proton spin density (M0) mapping to improve scan efficiency and reduce motion artifacts. To this end, a novel reconstruction was developed that combines low-rank subspace modeling with a deep image prior, termed DIP-MRF. A system of neural networks is used to generate spatial basis images and quantitative tissue property maps, with training performed using only the undersampled k-space measurements from the current scan. This approach avoids difficulties with obtaining in vivo MRF training data, as training is performed de novo for each acquisition. Calculation of the forward model during training is accelerated by using GRAPPA operator gridding to shift spiral k-space data to Cartesian grid points, and by using a neural network to rapidly generate fingerprints in place of a Bloch equation simulation. DIP-MRF was evaluated in simulations and at 1.5 T in a standardized phantom, 18 healthy subjects, and 10 patients with suspected cardiomyopathy. In addition to conventional mapping, two cardiac MRF sequences were acquired, one with a 15-heartbeat(HB) breathhold and 254 ms acquisition window, and one with a 5HB breathhold and 150 ms acquisition window. In simulations, DIP-MRF yielded decreased nRMSE compared to dictionary matching and a sparse and locally low rank (SLLR-MRF) reconstruction. Strong correlation (R2 > 0.999) with T1 and T2 reference values was observed in the phantom using the 5HB/150 ms scan with DIP-MRF. DIP-MRF provided better suppression of noise and aliasing artifacts in vivo, especially for the 5HB/150 ms scan, and lower intersubject and intrasubject variability compared to dictionary matching and SLLR-MRF. Furthermore, it yielded a better agreement between myocardial T1 and T2 from 15HB/254 ms and 5HB/150 ms MRF scans, with a bias of -9 ms for T1 and 2 ms for T2. In summary, this study introduces an extension of the deep image prior framework for cardiac MRF tissue property mapping, which does not require pre-training with in vivo scans, and has the potential to reduce motion artifacts by enabling a shortened breathhold and acquisition window.

Dataset Information

Deep learning to diagnose cardiac amyloidosis from cardiovascular magnetic resonance.

Publications

Deep learning to diagnose cardiac amyloidosis from cardiovascular magnetic resonance.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure