Dataset Information

Differentiation and quantification of fibrosis, fat and fatty fibrosis in human left atrial myocardium using ex vivo MRI.

ABSTRACT:

Background

Atrial fibrillation is associated with an atrial cardiomyopathy composed mainly of fibrosis and adipose tissue accumulation. We hypothesized that MRI, when used in an optimal ex vivo setting allowing high spatial resolution without motion artifacts, can help characterizing the complex 3D left atrial (LA) wall composition in human myocardial samples, as compared to histology.

Methods

This prospective case-control study was approved by the institutional review board. 3D MRI acquisitions including saturation-recovery T1 mapping and DIXON imaging was performed at 4.0 T on 9 human LA samples collected from patients who underwent cardiac surgery. Histological quantification of fibrosis and fat was obtained. MRI T1 maps were clustered based on a Gaussian Mixture Model allowing quantification of total, interstitial and fatty fibrosis components. Fat maps were computed from DIXON images and fat fractions were calculated. MRI measurements were performed on the same location as the histological analysis (plane) and on the entire sample volume (3D).

Results

High correlations and levels of agreement were observed between MRI and histology for total (r = 0.93), interstitial (r = 0.93) and fatty fibrosis (r = 0.98) and fat (r = 0.96). Native T1 correlated with the amount of fibrosis from MRI and histology. The 3D MRI total, interstitial and fatty fibrosis ranges were between 6% and 23%, 4% and 17.3%; and 1.4% and 19.7% respectively.

Conclusion

High Field ex vivo MRI was able to quantify different LA myocardial components with high agreement in 2D with histology and moreover to provide 3D quantification of such components whereas in vivo application remains a challenge.

SUBMITTER: Bouazizi K

PROVIDER: S-EPMC6175289 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Differentiation and quantification of fibrosis, fat and fatty fibrosis in human left atrial myocardium using ex vivo MRI.

Bouazizi Khaoula K Rahhal Amer A Kusmia Slawomir S Evin Morgane M Defrance Carine C Cluzel Philippe P Berthet Myriam M Atassi Fabrice F Leprince Pascal P Lebreton Guillaume G Kachenoura Nadjia N Hatem Stéphane N SN Redheuil Alban A

PloS one 20181008 10

<h4>Background</h4>Atrial fibrillation is associated with an atrial cardiomyopathy composed mainly of fibrosis and adipose tissue accumulation. We hypothesized that MRI, when used in an optimal ex vivo setting allowing high spatial resolution without motion artifacts, can help characterizing the complex 3D left atrial (LA) wall composition in human myocardial samples, as compared to histology.<h4>Methods</h4>This prospective case-control study was approved by the institutional review board. 3D M ...[more]

PMID: 30296279

Similar Datasets

Project description:BackgroundEpicardial adipose tissue (EAT) may induce left atrium (LA) wall inflammation and promote LA fibrosis. Therefore, simultaneous assessment of these two important atrial fibrillation (AF) risk factors would be desirable.PurposeTo perform a comprehensive evaluation of 3D Dixon water-fat separated late gadolinium enhancement (LGE-Dixon) MRI by analysis of repeatability and systematic comparison with reference methods for assessment of fibrosis and fat.Study typeProspective.PopulationTwenty-eight, 10, and 7 patients, respectively, with clinical indications for cardiac MRI.Field strength/sequenceA 1.5-T scanner, inversion recovery multiecho spoiled gradient echo.AssessmentTwenty-eight patients (age 58 ± 19 years, 15 males) were scanned using LGE-Dixon. A 5-point Likert-type scale was used to grade the image quality. Another 10 patients (age 46 ± 19 years, 9 males) were scanned using LGE-Dixon and 3D proton density Dixon (PD-Dixon). Finally, seven patients (age 62 ± 14 years, 4 males) were scanned using LGE-Dixon and conventional LGE. The scan time, intraobserver and interobserver variability, and levels of agreement were assessed.Statistical testsStudent's t-test, one-way ANOVA, and Mann-Whitney U-test were used; P < 0.05 was considered significant, intraclass correlation coefficient (ICC).ResultsThe scan time (minutes:seconds) for LGE-Dixon (n = 28) was 5:01 ± 1:40. ICC values for intraobserver and interobserver measurements of LA wall fibrosis percentage were 0.98 (95% CI, 0.97-0.99) and 0.97 (95% CI, 0.94-0.99) while of EAT were 0.92 (95% CI, 0.82-0.97) and 0.90 (95% CI, 0.80-0.95). The agreement for LA fibrosis percentage between the LGE-Dixon and the conventional LGE was 0.92 (95% CI, 0.66-0.99) and for EAT volume between the LGE-Dixon and the PD-Dixon was 0.93 (95% CI, 0.72-0.98).ConclusionLA fibrosis and EAT can be assessed simultaneously using LGE-Dixon. This method allows a high level of intraobserver and interobserver repeatability as well as agreement with reference methods and can be performed in a clinically feasible scan time.Evidence level2 TECHNICAL EFFICACY STAGE: 3.

Project description:PurposeTo assess the performance of an automated myocardial T2 and extracellular volume (ECV) quantification method using transfer learning of a fully convolutional neural network (CNN) pretrained to segment the myocardium on T1 mapping images.Materials and methodsA single CNN previously trained and tested using 11 550 manually segmented native T1-weighted images was used to segment the myocardium for automated myocardial T2 and ECV quantification. Reference measurements from 1525 manually processed T2 maps and 1525 ECV maps (from 305 patients) were used to evaluate the performance of the pretrained network. Correlation coefficient (R) and Bland-Altman analysis were used to assess agreement between automated and reference values on per-patient, per-slice, and per-segment analyses. Furthermore, transfer learning effectiveness in the CNN was evaluated by comparing its performance to four CNNs trained using manually segmented T2-weighted and postcontrast T1-weighted images and initialized using random-weights or weights of the pretrained CNN.ResultsT2 and ECV measurements using the pretrained CNN strongly correlated with reference values in per-patient (T2: R = 0.88, 95% confidence interval [CI]: 0.85, 0.91; ECV: R = 0.91, 95% CI: 0.89, 0.93), per-slice (T2: R = 0.83, 95% CI: 0.81, 0.85; ECV: R = 0.84, 95% CI: 0.82, 0.86), and per-segment (T2: R = 0.75, 95% CI: 0.74, 0.77; ECV: R = 0.76, 95% CI: 0.75, 0.77) analyses. In Bland-Altman analysis, the automatic and reference values were in good agreement in per-patient (T2: 0.3 msec ± 2.9; ECV: -0.3% ± 1.7), per-slice (T2: 0.1 msec ± 4.6; ECV: -0.3% ± 2.5), and per-segment (T2: 0.0 msec ± 6.5; ECV: -0.4% ± 3.5) analyses. The performance of the pretrained network was comparable to networks refined or trained from scratch using additional manually segmented images.ConclusionTransfer learning extends the utility of pretrained CNN-based automated native T1 mapping analysis to T2 and ECV mapping without compromising performance. Supplemental material is available for this article. © RSNA, 2020.

Dataset Information

Differentiation and quantification of fibrosis, fat and fatty fibrosis in human left atrial myocardium using ex vivo MRI.

Background

Methods

Results

Conclusion

Publications

Differentiation and quantification of fibrosis, fat and fatty fibrosis in human left atrial myocardium using ex vivo MRI.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets