A token-mixer architecture for CAD-RADS classification of coronary stenosis on multiplanar reconstruction CT images.

BACKGROUND AND OBJECTIVE: In patients with suspected Coronary Artery Disease (CAD), the severity of stenosis needs to be assessed for precise clinical management. An automatic deep learning-based algorithm to classify coronary stenosis lesions according to the Coronary Artery Disease Reporting and Data System (CAD-RADS) in multiplanar reconstruction images acquired with Coronary Computed Tomography Angiography (CCTA) is proposed. METHODS: In this retrospective study, 288 patients with suspected CAD who underwent CCTA scans were included. To model long-range semantic information, which is needed to identify and classify stenosis with challenging appearance, we adopted a token-mixer architecture (ConvMixer), which can learn structural relationship over the whole coronary artery. ConvMixer consists of a patch embedding layer followed by repeated convolutional blocks to enable the algorithm to learn long-range dependences between pixels. To visually assess ConvMixer performance, Gradient-Weighted Class Activation Mapping (Grad-CAM) analysis was used. RESULTS: Experimental results using 5-fold cross-validation showed that our ConvMixer can classify significant coronary artery stenosis (i.e., stenosis with luminal narrowing ≥50%) with accuracy and sensitivity of 87% and 90%, respectively. For CAD-RADS 0 vs. 1-2 vs. 3-4 vs. 5 classification, ConvMixer achieved accuracy and sensitivity of 72% and 75%, respectively. Additional experiments showed that ConvMixer achieved a better trade-off between performance and complexity compared to pyramid-shaped convolutional neural networks. CONCLUSIONS: Our algorithm might provide clinicians with decision support, potentially reducing the interobserver variability for coronary artery stenosis evaluation.

Emerging Role of Cardiac Magnetic Resonance Imaging in Diagnosing Myocarditis: A Blunder or The Way To Get the Problem Sorted?

Acute myocarditis is a disease affecting the myocardial tissue, which is caused by infections, rheumatic diseases, especially sarcoidosis, or certain therapies. Its diagnosis may be difficult, owing to its variable clinical presentation. In this setting, cardiac magnetic resonance plays a pivotal role in detecting myocardial inflammation through qualitative, semiquantitative, and quantitative parameters, in particular with the new quantitative techniques such as T1 and T2 mapping, combined or not with late gadolinium enhancement evaluation. This is in accordance with the revised Lake Louise criteria. In this review, the emerging role of the new cutting-edge cardiac magnetic resonance imaging techniques in diagnosing myocarditis is extensively presented.

Could CMR Tissue-Tracking and Parametric Mapping Distinguish Between Takotsubo Syndrome and Acute Myocarditis? A Pilot Study.

RATIONALE AND OBJECTIVE: Takotsubo syndrome (TS) is a transient and often misdiagnosed form of left ventricular dysfunction. Acute myocarditis (AM) is usually included in TS differential diagnosis. The aim of this study is to assess the role of cardiac magnetic resonance imaging coupled with tissue-tracking technique (CMR-TT) and parametric mappings analysis in discriminating between TS and AM. MATERIALS AND METHODS: We retrospectively enrolled three groups: patients with TS (n = 12), patients with AM (n = 14), and 10 healthy controls. All the patients had a comprehensive CMR examination, including the assessment of global and segmental longitudinal strain, circumferential strain, radial strain (RS), and parametric mapping. RESULTS: The analysis of variance was used to compare the different groups. In TS patients, basal RS, global T1 mapping, global T2 mapping, mid T2 mapping, apical T1 and T2 mapping were statistically significantly different compared with the other groups. MANCOVA analysis confirmed that the association between myocardial strain data and parametric mapping was independent on age and sex. Apical T1 and T2 mapping proved to have a good performance in differentiating TS from AM (area under the curves of 0.908 and 0.879, respectively). CONCLUSION: Basal RS and apical tissue mapping analysis are the most advanced CMR-derived parameters in making a differential diagnosis between TS and AM.

Atrial Strain by Feature-Tracking Cardiac Magnetic Resonance Imaging in Takotsubo Cardiomyopathy. Features, Feasibility, and Reproducibility.

OBJECTIVES: The purpose of this study was to investigate whether there may be a bi-atrial dysfunction in Takotsubo syndrome (TS) during the transient course of the disease, using cardiac magnetic resonance imaging feature tracking (CMR-FT) in analyzing bi-atrial strain. METHOD: Eighteen TS patients and 13 healthy controls were studied. Reservoir, conduit, and booster bi-atrial functions were analyzed by CMR-FT. The correlation between LA and RA strain parameters was assessed. Intra- and inter-observer reproducibility was evaluated for all strain and strain rate (SR) parameters using intraclass correlation coefficients (ICCs) and Bland-Altman analysis. RESULTS: Atrial strain were feasible in all patients and controls. Takotsubo patients showed an impaired LA Reservoir strain (∊s), LA Reservoir strain rate (SRs), LA and RA Conduit strain(∊e), LA and RA conduit strain rate (SRe) in comparison with controls (P < 0.001 for all of them), while no differences were found as to LA and RA booster deformation parameters (∊a and SRa). Analysis of correlation showed that LA ∊s, SRs, ∊e, and SRe were positively correlated with corresponding RA strain measurements (P < 0.001, r = 0.61 and P = 0,03, r = 0,54, respectively). Reproducibility was good to excellent for all atrial strain and strain rate parameters (ICCs ranging from 0,50 to 0,96). CONCLUSION: Atrial strain analysis using CMR-FT may be a useful tool to reveal new pathophysiological insights in Takotsubo cardiomyopathy. Additional studies, with a larger number of patients, are needed to confirm the possible role of these advanced CMR tools in characterizing TS patients.

Anderson-fabry disease: role of traditional and new cardiac MRI techniques.

Anderson-Fabry (FD) disease is a rare X-linked disorder caused by different mutations in the Galactosidase α (GLA) gene, which leads to α-galactosidase A enzyme deficiency and the storage of glycosphingolipids in different kinds of organs, included the heart. This results in myocardial inflammation and left ventricular hypertrophy (LVH) and fibrosis. Echocardiography and cardiac magnetic resonance (C-MRI), in particular with new techniques, such as mapping analysis, late gadolinium enhancement (LGE) assessment and strain imaging, are important tools that allow a correct diagnosis, discriminating FD from other hypertrophic heart conditions. C-MRI is able to detect tissue alterations in the early stages of the disease, when an appropriate treatment could be more effective, and it has a fundamental role in monitoring therapy.