Myocardial extracellular volume fraction quantification based on T1 mapping at 3 T: quality optimization by contour-based registration and segmental analysis

Abstract

Background

Myocardial extracellular volume fraction (ECV) assessment can be affected by various technical and subject-related factors.

Purpose

To evaluate the role of contour-based registration in quantification of ECV and investigate normal segment-based myocardial ECV values at 3T.

Material and Methods

Pre- and post-contrast T1 mapping images of the left ventricular basal, mid-cavity, and apical slices were obtained in 26 healthy volunteers. ECV maps were generated using motion correction with and without contour-based registration. The image quality of all ECV maps was evaluated by a 4-point scale. Slices were dichotomized according to the occurrence of misregistration in the source data. Contour-registered ECVs and standard ECVs were compared within each subgroup using analysis of variance for repeated measurements and generalized linear mixed models.

Results

In all three slices, higher quality of ECV maps were found using contour-registered method than using standard method. Standard ECVs were statistically different from contour-registered ECVs in global (26.8% ± 2.8% vs. 25.8% ± 2.4%; P = 0.001), mid-cavity (25.4% ± 3.1% vs. 24.3% ± 2.5%; P = 0.016), and apical slices (28.7% ± 4.1% vs. 27.2% ± 3.4%; P = 0.010). In the misregistration subgroups, contour-registered ECVs were lower with smaller SDs (basal: 25.2% ± 1.8% vs. 26.7% ± 2.6%; P = 0.038; mid-cavity: 24.4% ± 2.3% vs. 26.8% ± 3.1%; P = 0.012; apical: 27.5% ± 3.6% vs. 29.7% ± 4.5%; P = 0.016). Apical (27.2% ± 3.4%) and basal-septal ECVs (25.6% ± 2.6%) were statistically higher than mid-cavity ECV (24.3% ± 2.5%; both P < 0.001).

Conclusion

Contour-based registration can optimize image quality and improve the precision of ECV quantification in cases demonstrating ventricular misregistration among source images.

Keywords

Cardiac magnetic resonance imaging artifacts

Introduction

Myocardial extracellular volume fraction (ECV) has been considered as a promising biomarker for the diagnosis and prognosis of various cardiac diseases due to its high sensitivity, reproducibility, and intrinsically quantitative nature (1). Nevertheless, T1 mapping techniques can be affected by diverse technical and subject-related factors (2,3). Significant heterogeneity was found among ranges of ECV in healthy controls in previous studies (4).

It is important for T1/ECV map generation that the images involved are anatomically aligned with each other so that pixel-specific calculations are correctly done on corresponding anatomical positions. Various automatic motion correction and co-registration approaches have been developed to suppress motion artifacts in T1/ECV maps (5 –14). However, fully automatic approaches are limited in several aspects related to the algorithms: intensity-based data-driven registration approaches could be complicated by contrast inversion, partial volume effect, and signal nulling for images near the zero crossing of the T1-relaxation curve (5,7 –11); model-based approaches depend on an accurate initialization for T1 and a sufficiently trained model, which might not be easily achieved in practice (5,12); the automatic segmentation-based approach only enforces registration at the endo- and epicardium, which limits its use in focal diseases (13). To overcome these limitations, a contour-based registration approach has been developed to align serial images using manually predefined myocardial contours, through polynomial transform and pixel interpolation within the myocardium. It remains unclear whether the contour-based registration has added values in ECV evaluation in practice.

In most previous studies, only mid-cavity ECV was analyzed due to good reproducibility and the limitation of postprocessing technique (15 –18). Studies reporting ECV values on a per-segment or per-slice basis are limited with controversial results (16,19 –23).

In this study, we hypothesized that contour-based registration can further optimize the image quality of ECV maps and calculation of ECV values on the basis of intensity-based automatic registration (5). We also aimed to investigate the distribution of myocardial ECV on the basis of American Heart Association (AHA) segmentation, as well as among different slices.

Material and Methods

Study population

The present study was approved by the institutional ethics committee. Written informed consent was obtained from all the volunteers. In total, 26 adult volunteers (16 men, 10 women; age range = 24–65 years) were enrolled from the health check-up centers in collaboration with our hospital. Blood tests were taken as part of the annual health check-ups of the participants.

The inclusion criteria were as follows: (i) normal blood pressure and normal blood biochemistry results; (ii) normal 12-lead electrocardiogram; and (iii) no history or symptoms of cardiovascular diseases, hypertension, nephropathy, thyroid diseases, or any other systemic diseases.

The exclusion criteria were as follows: (i) contraindications of cardiovascular magnetic resonance (CMR) examination; (ii) pregnancy, lactation, or preparing for pregnancy; (iii) allergic to gadolinium contrast agent; (iv) incomplete clinical data or insufficient compliance with CMR examinations.

CMR protocols

All CMR examinations were performed on a 3-T scanner (Magnetom Verio; Siemens Healthcare, Erlangen, Germany). CMR protocols included dark blood imaging, cine imaging, pre- and post-contrast T1 mapping and late gadolinium enhancement (LGE) imaging. Pre-contrast T1-mapping was performed using an electrocardiogram-gated Modified Look-Locker Inversion Recovery (MOLLI) with a 5(3)3 strategy (17), while post-contrast T1 mapping was performed 15 min after the administration of 0.2 mmol/kg gadoterate meglumine (Beilu, Beijing, PR China) with a 4(1)3(1)2 strategy (24). Both pre- and post-contrast T1-mapping images were obtained in the basal, mid-cavity, and apical short-axis slices, with the following parameters: repetition time/echo time = 2.3/1.1 ms; flip angle = 35°; field of view = 306 × 360 mm; matrix = 218 × 256; slice thickness = 8 mm; bandwidth = 1085 Hz/Px; minimum inversion time = 120 ms; inversion time increment = 80 ms; 7/8 partial Fourier plus parallel acquisition technique with acceleration factor of 2. Images at different inversion time were automatically motion-corrected and T1 maps (hereafter referred to as standard T1 maps) were generated on the scanner right after each scan (5).

Image analysis

General analysis

Left ventricular functions of all the participants were calculated (Argus; Siemens Medical Solutions, Erlangen, Germany) and all the images were reviewed by two radiologists (one with three years and the other with more than ten years of experience in CMR) separately to identify any incidental cardiovascular diseases, such as valve disease, congenital cardiovascular diseases or myocardial late gadolinium enhancement, and so on.

Contour-based registration

The contour-based registration is a feature-based registration developed by CVI42 (v5.0, Circle Cardiovascular Imaging, Alberta, Canada). This registration method uses control points to map images to a selected reference image, and to align the anatomical features within the myocardial contours. Corresponding landmark points of the pre- and post-contrast T1 mapping images are collected from the short-axis insertion points, and the drawn endo- and epi-myocardial contours. The system of the two sets of landmark points is set up using polynomials, and then solved to acquire a transform that best maps the set of points from one image to the other. The calculated transform is used to apply inverse mapping from the reference image to the second image, thus allowing the mapping of image intensities from the second image to the corresponding locations that are aligned with the reference image; image intensities for transformed non-discrete indices are estimated from surrounding pixels using cubic interpolation. In this way, every image in the series is registered with the selected reference image. This then creates two new sets of pre- and post- T1 images where the myocardial anatomy is aligned with the reference image. Once the myocardium is aligned throughout all the images, two new pixel-wise T1 maps are generated, and were referred to as contour-registered T1 maps (Fig. 1). A screenshot of the contour-based registration in the software can be found in Supplement S1.

Fig. 1.

The process of contour-based registration. The epi- and endo-myocardial contours of the selected pre-contrast T1-mapping image (yellow frame) were copied and pasted to all the other pre- and post-contrast images. Part of the contours did not fit well into the posterior wall and the septum in two post-contrast images (red frames and white arrows). After contour-based registration, the contour-registered T1 maps were generated. The artifacts in the standard ECV map were not visible in the contour-registered ECV map. ECV, extracellular volume fraction.

Calculation of ECV

Hematocrit (HCT) was obtained from the blood sample collected on the same day of the CMR examination for each participant. Pre- and post-contrast T1 values of the blood pool were collected by drawing regions of interest in the left ventricle. Pixel-wise standard ECV maps and contour-registered ECV maps were generated by the software according to the previously reported equation (25) (Fig. 1).

Evaluation for the effect of the contour-based registration

The image quality of all ECV maps were reviewed and graded as “excellent,” “good,” “fair,” or “poor” by two radiologists who did not participate in the contour-based registration processes. Excellent ECV maps were those demonstrating homogenous myocardial signal intensity, with clear margin of myocardium and no artifacts. Good maps were those with slightly blurred margin of myocardium and/or minor artifacts affecting <50% of the mural thickness. Fair images were those presenting obvious artifacts affecting >50% of the mural thickness, but less than half of the segments were affected. Poor quality meant more than half of the segments were affected by severe artifacts (Fig. 2).

Fig. 2.

Grading for image quality of the ECV maps. A = excellent; B = good; C = fair; D = poor. ECV, extracellular volume fraction.

Both standard and contour-registered ECV values of each slice and AHA segment (26) were obtained automatically by the software. The global ECV was calculated by averaging ECVs of the three slices, excluding segments affected by severe artifacts.

Misregistration of left ventricle was identified when the reference contour did not fit with one or more images of the pre- or post-contrast T1 images or T1maps (Fig. 1). There were two types of misregistration errors. Type I was a misregistration between pre- and post-contrast T1 maps. Type II was a misregistration of T1 source images of different inversion times. Participants were divided into two subgroups for each short-axis slice based on the existence of left ventricular misregistration. Comparisons of standard ECVs and contour-registered ECVs were performed within each subgroup.

Global ECVs and ECVs of the three short-axis slices of all the participants were compared. ECVs of the 16 segments were also compared. All the analyses were performed using standard ECVs and contour-registered ECVs, respectively.

Statistical analysis

Statistical analyses were performed using SPSS version 23.0 (IBM Corp. Armonk, NY, USA). T1 mapping images of 10 randomly selected participants were contoured again one month later by the same radiologist to evaluate the intra-observer agreement using Bland–Altman plots and interclass correlation coefficient (ICC). Paired-samples t-test was used to compare the contour-registered ECVs and standard ECVs. One-way analysis of variance (ANOVA) for repeated measures was applied for comparisons among ECVs of different slices and layers. For multiple comparisons among ECVs of 16 segments, generalized linear mixed model was applied to correct repeated measures. P < 0.05 was considered significant.

Results

The demographic characteristics and left ventricular function of the all the participants were all within the normal range (Table 1). The basal ECV map of one participant was not available because the basal post-contrast T1 images were lost due to a technical failure. Thus, T1 maps of 25 basal slices, 26 mid-cavity slices, and 26 apical slices were taken into analysis. ICC indicated satisfactory intra-observer reproducibility of the contour-based registration, which was 0.966 (95% confidence interval [CI] = 0.908–0.984). The mean difference in Bland–Altman test was −0.32% (95% CI = −1.29 to 0.65).

Table 1.

Demographic characteristics and left ventricular function.

Characteristics
Age (years)	46.1 ± 10.6
Height (cm)	167.0 ± 8.5
Weight (kg)	69.3 ± 13.1
BMI (kg/m²)	24.7 ± 3.1
Body surface area (m²)	1.88 ± 0.20
Heart rate (/min)	69.1 ± 7.7
Blood pressure (systolic/diastolic) (mmHg)	118.2 ± 13.1/73.4 ± 9.5
End-diastole volume (mL)	140.1 ± 38.8
End-systole volume (mL)	55.5 ± 22.8
Ejection fraction (%)	61.1 ± 7.3
Cardiac output (L/min)	5.83 ± 1.73
End-diastole myocardial mass (g)	97.5 ± 36.5
End-diastole myocardial mass / body surface area (g/m²)	54.5 ± 21.8

Values are given as mean ± SD.

BMI, body mass index; SD, standard deviation.

Effect of contour-based registration

Grading for image quality of the standard and contour-registered ECV maps are shown in Table 2. In all three slices, there were larger proportions of excellent and good contour-registered ECV maps than those of the standard ECV maps.

Table 2.

Number of participants with different image quality grades of the ECV maps.

	Basal slice (n = 25)		Mid-cavity slice (n = 26)		Apical slice (n = 26)
Grade	Standard	Contour-registered	Standard	Contour-registered	Standard	Contour-registered
Excellent	11 (44)	15 (60)	10 (39)	18 (69)	4 (15)	7 (27)
Good	10 (40)	8 (32)	12 (46)	7 (27)	10 (39)	10 (39)
Fair	2 (8)	1 (4)	3 (12)	1 (4)	11 (42)	9 (35)
Poor	2 (8)	1 (4)	1 (4)	0	1 (3.8)	0

Values are given as n (%).

ECV, extracellular volume fraction.

For 25% (19/77) of the slices, contour-based registration improved the image quality with changes in grading. For the other slices, despite the same quality grades of standard ECV and contour-registered ECV maps, many of the latter demonstrated a clearer margin of the myocardium with less artifacts or more homogeneous signal intensity. There was no decrease of image quality after contour-based registration.

A total of 16 participants demonstrated ventricular misregistration in one or more slices. Type II misregistration error was more prevalent than Type I (Table 3). In most participants with ventricular misregistration, the minor artifacts demonstrated in standard ECV maps disappeared or shrunk in contour-registered ECV maps (Table 3, Figs. 1,3–5). In non-misregistration subgroups, the image quality of standard ECV was consistent with contour-registered ECV maps (Fig. 6).

Fig. 3.

(a–c) Standard and (d–f) contour-registered T1/ECV maps of the basal slice of a 55-year-old woman. The artifacts (c, white arrow) caused by inconsistent ventricular shape between pre- and post-contrast T1 maps were mitigated after contour-based registration (f). ECV, extracellular volume fraction.

Fig. 4.

(a–c) Standard and (d–f) contour-registered T1/ECV maps of the mid-cavity slice of a 41-year-old man. (c) The artifacts in the lateral wall caused by inconsistent ventricular shape among post-contrast T1 images (the contours did not fit into the image in the red frame) were eliminated after contour-based registration (f). ECV, extracellular volume fraction.

Fig. 5.

(a–c) Standard and (d–f) contour-registered T1/ECV maps of the apical slice of a 39-year-old man. (c) The artifacts in the posterior wall caused by inconsistent ventricular shape among post-contrast T1 images (the contours did not fit into the image in the red frame) were eliminated after contour-based registration (f). ECV, extracellular volume fraction.

Fig. 6.

(a–c) Standard and (d–f) contour-registered T1/ECV maps of the mid-cavity slice of a 34-year-old man in the non-misregistration subgroup. There was no observable difference between (c) the standard ECV map and (f) the contour-registered ECV map. ECV, extracellular volume fraction.

Table 3.

Left ventricular misregistration in three short-axis slices and the changes in ECV maps after contour-based registration.

Ventricular misregistration	Basal slice n = 25		Mid-cavity slice n = 26		Apical slice n = 26
Ventricular misregistration	Type I	Type II	Type I	Type II	Type I	Type II
No. of participants	4 (16)	7 (28)	4 (15)	7 (27)	7 (27)	8 (31)
Participants with artifacts disappeared or shrunk*	3 (12)	6 (24)	3 (12)	6 (23)	7 (27)	4 (15)
Participants with elevated image grades*	2 (8)	4 (16)	3 (12)	6 (23)	3 (12)	4 (15)

Values are given as n (%).

*Contour-registered ECV map compared with standard ECV map.

ECV, extracellular volume fraction.

As for ECV values, 49 of 410 segments (12%) were excluded from the calculation of standard global ECVs due to obvious artifacts. After contour-based registration, only 29 of 410 segments (7%) remained excluded from the calculation of contour-registered global ECVs. Two basal standard ECV maps with “poor” grading were excluded from statistical analysis because all the segments were affected by obvious artifacts. Table 4 demonstrates that excluding the basal slices, standard ECVs were still statistically higher than contour-registered ECVs in global (0.001), mid-cavity (0.016), and apical slices (0.010).

Table 4.

Paired-samples comparisons between standard ECVs and contour-registered ECVs.

	Standard ECV (%)	Contour-registered ECV (%)	t value	P value
Global	26.8 ± 2.8	25.8 ± 2.4	3.89	0.001
Basal	26.2 ± 2.7	25.6 ± 2.7	1.86	0.077
Mid-cavity	25.4 ± 3.1	24.3 ± 2.6	2.58	0.016
Apical	28.7 ± 4.1	27.2 ± 3.4	2.79	0.010

Values are given as mean ± SD unless otherwise indicated.

ECV, extracellular volume fraction.

The contour-registered ECVs were all lower than the standard ECVs with smaller SDs in all misregistration subgroups (basal P = 0.038; mid-cavity P = 0.012; apical P = 0.016). There were no statistical differences between contour-registered ECVs and standard ECVs in non-misregistration subgroups of all the three slices (basal P = 0.518; mid-cavity P = 0.707; apical P = 0.253) (Table 5).

Table 5.

Paired-samples comparisons between contour-registered ECVs and standard ECVs in different subgroups.

Slice	Subgroup		ECV (%)	P value
Basal	Misregistration (n = 11, 44%)	Standard	26.7 ± 2.6	0.038
	Misregistration (n = 11, 44%)	Contour-registered	25.2 ± 1.8	0.038
	Non-misregistration (n = 12, 56%)	Standard	25.8 ± 2.9	0.518
	Non-misregistration (n = 12, 56%)	Contour-registered	25.9 ± 3.4	0.518
Mid-cavity	Misregistration (n = 11, 42%)	Standard	26.8 ± 3.1	0.012
	Misregistration (n = 11, 42%)	Contour-registered	24.4 ± 2.3	0.012
	Non-misregistration (n = 15, 58%)	Standard	24.3 ± 2.6	0.707
	Non-misregistration (n = 15, 58%)	Contour-registered	24.3 ± 2.7	0.707
Apical	Misregistration (n = 15, 58%)	Standard	29.7 ± 4.5	0.016
	Misregistration (n = 15, 58%)	Contour-registered	27.5 ± 3.6	0.016
	Non-misregistration (n = 11, 42%)	Standard	27.2 ± 3.2	0.253
	Non-misregistration (n = 11, 42%)	Contour-registered	26.9 ± 3.2	0.253

Values are given as mean ± SD unless otherwise indicated.

ECV, extracellular volume fraction.

For both standard ECVs and contour-registered ECVs, the apical ECVs were higher with larger SDs than those of the other two slices (Table 6). Generalized linear mixed models showed that ECVs of the 16 segments were not all equal (P < 0.001 for both standard ECVs and contour-registered ECVs). Pairwise comparisons with Bonferroni correction for standard ECVs and contour-registered ECVs, respectively, indicated that ECVs of segments 2, 3, 13, 14, and 16 were statistically higher than those of the other segments (Supplement S2).

Table 6.

Multiple comparisons among global ECVs and ECVs of different slices using one-way ANOVA for repeated measures.

	Myocardial ECV (%)
Method	Global	Basal	Mid-cavity	Apical	F	P
Standard	26.8 ± 2.8	26.2 ± 2.7	25.4 ± 3.1	28.7 ± 4.1	11.4	<0.001
Contour-registered	25.8 ± 2.4	25.6 ± 2.6	24.3 ± 2.5	27.2 ± 3.4	12.7	<0.001

Values are given as mean ± SD.

ANOVA, analysis of variance; ECV, extracellular volume fraction; SD, standard deviation.

Discussion

In the present study, contour-based registration optimized image quality of T1/ECV maps and improved the precision of ECV quantification, with particular effect on ventricular misregistration. Higher ECV values were found in apical slice and basal-septal segments than the rest of the myocardium even after motion correction and contour-based registration.

In this study, image quality of ECV maps visibly improved with statistically significant decrease of ECV values after contour-based registration. The percentage of excellent and good standard ECV maps (85%) of the mid-cavity slice was similar with previously reported results (83%) using the same automatic approach (6). After contour-based registration, it rose up to 96%. Two aspects of improvement were observed in misregistration subgroups. ECV values after contour-based registration were lower with smaller SDs, which was consistent with fewer artifacts and more homogeneous signal intensities in contour-registered ECV maps. As Table 5 shows, the statistical difference between contour-registered ECVs and standard ECVs were mainly caused by ventricular misregistration. Our findings suggested that contour-based registration could optimize image quality of T1/ECV maps and improve the precision of ECV quantification in the misregistration subgroups, while no impact on the non-misregistration subgroups was observed.

The incidence of left ventricular misregistration during T1 mapping scan was higher than expected in this study. All the volunteers had normal routine 12-lead ECGs and were capable of steady breath-hold. Yet, the misregistration subgroup accounted for 44% of the participants in basal slice, 42% in mid-cavity slice, and 58% in apical slice. Although the majority merely demonstrated minor misregistration affecting limited segments in one or several images or between pre- and post-contrast T1 maps, the improvement of image quality after contour-based registration was evident and there was statistically significant change in ECV values even after excluding the segments affected by artifacts. The mitigation of the artifacts in specific segments might be essential for evaluation of focal lesions.

Previous studies suggested that subject-related factors such as heart rate, breath-hold, and arrhythmia could affect image quality of T1 mapping technique. In this study, R-R interval change and different depth of breath-hold were two main factors that caught our attention. R-R interval change was common when people hold their breath. This physiological phenomenon might be so evident in certain individuals that it led to the misregistration in one or several images among T1 images at different inversion time. On the other hand, the inconsistency of the pre- and post-contrast T1 maps was mainly caused by different depth of breath-holds between scans. In addition, occasional arrhythmia during a scan such as ventricular extra systole might also contribute to ventricular misregistration, which was hard to be revealed by routine ECG. We suppose that misregistration would be more prevalent in patients with cardiac and pulmonary conditions, who would have higher prevalence of arrhythmia and less capacity of breath-hold. While automatic motion correction is fast and effective for clinical routine, the contour-based registration could serve as a complement for the fully automated clinical workflow. Its potential value is promising in individualized medicine and in research field.

Global myocardial ECVs in this study (standard ECV: 26.8% ± 2.8%, contour-registered ECV: 25.8% ± 2.4%) were similar to the results in previous studies with 3T modality (4,16,19,20). However, the apical slice and basal septum demonstrated higher ECV than the other slices and segments. This result conflicted with one previous study that demonstrated similar ECVs in all slices and segments (20). The apical short-axis slice had thinner myocardium and more motion artifacts than the basal and mid-cavity slices. In this study, the apical ECV maps presented the lowest image quality and the highest prevalence of ventricular misregistration (58%) among all the three slices. Both standard ECVs and contour-registered ECVs of the apical slice were higher with larger SDs than those of the other two slices, indicating limited effects of motion-correction and contour-based registration on apical slice. As the basal slice was localized near the left ventricular outflow tract, segments 2 and 3 were thinner than the other segments in the same slice and were more susceptible to partial volume effect and cardiac motion. Therefore, the regional variance of myocardial ECVs was more likely due to technical limitations rather than true tissue composition.

Considering the increasing use of T1 mapping technique for focal myocardial lesions, normal ECV ranges specifically for apical slice and basal-septal segments need to be considered before algorithms capable of eliminating residual motion are developed.

This study is limited by the small sample size. However, since this is a self-contrast design and regional ECV was analyzed for slices and segments, the sample size is sufficient to test our hypothesis. Another limitation is that we only evaluated the effect of contour-based registration on T1 mapping images obtained by MOLLI sequence. Therefore, our findings may not be expanded to other T1 mapping sequences such as shMOLLI, SASHA, and SAPPHIRE.

In conclusion, contour-based registration can optimize image quality of T1/ECV maps in cases with misregistration among source images, thus leading to higher precision of ECV quantification. Myocardial ECVs of the apical slice and basal-septal segments were higher than other slices and segments, probably due to technical limitations rather than true tissue composition. Clinical myocardial tissue characterization regarding the above areas should be interpreted with caution. Region-specific normal ranges of ECV might be needed for studies evaluating focal lesions.

Supplemental Material

sj-docx-1-acr-10.1177_02841851211067149 - Supplemental material for Myocardial extracellular volume fraction quantification based on T1 mapping at 3 T: quality optimization by contour-based registration and segmental analysis

Supplemental material, sj-docx-1-acr-10.1177_02841851211067149 for Myocardial extracellular volume fraction quantification based on T1 mapping at 3 T: quality optimization by contour-based registration and segmental analysis by Hong Dai, YutaoWang, Randi Fu, Sijia Ye, Xiuchao He, Shuying Luo and Wei Jin in Acta Radiologica

Footnotes

Acknowledgements

The authors thank Professor Sheng-Xian Tu from Shanghai Jiao Tong University for his advice in methodology and manuscript organization.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This study was funded by Guangdong Science and Technology Department (2013B021800136). Ling Lin was supported by the China Scholarship Council (CSC201807720065).

ORCID iDs

Ling Lin

Xu-Hui Zhou

Supplemental material

Supplemental material for this article is available online.

References

Schelbert

Messroghli

. State of the art: clinical applications of cardiac T1 mapping. Radiology 2016;278:658–676.

Moon

Messroghli

Kellman

, et al. Myocardial T1 mapping and extracellular volume quantification: a Society for Cardiovascular Magnetic Resonance (SCMR) and CMR working group of the European Society of Cardiology consensus statement. J Cardiovasc Magn Reson 2013;15:92.

Messroghli

Moon

Ferreira

, et al. Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: a consensus statement by the Society for Cardiovascular Magnetic Resonance (SCMR) endorsed by the European Association for Cardiovascular Imaging (EACVI). J Cardiovasc Magn Reson 2017;19:75.

Gottbrecht

Kramer

Salerno

. Native T1 and extracellular volume measurements by cardiac MRI in healthy adults: a meta-analysis. Radiology 2019;290:317–326.

Xue

Shah

Greiser

, et al. Motion correction for myocardial T1 mapping using image registration with synthetic image estimation. Magn Reson Med 2012;67:1644–1655.

Kellman

Wilson

Xue

, et al. Extracellular volume fraction mapping in the myocardium, part 1: evaluation of an automated method. J Cardiovasc Magn Reson 2012;14:63.

Roujol

Foppa

Weingärtner

, et al. Adaptive registration of varying contrast-weighted images for improved tissue characterization (ARCTIC): application to T1 mapping. Magn Reson Med 2015;73:1469–1482.

Huizinga

Poot

Guyader

J-M

, et al. PCA-based groupwise image registration for quantitative MRI. Med Image Anal 2016;29:65–78.

Song

Stainsby

Hood

Slavin

. Flexible cardiac T1 mapping using a modified Look -Locker acquisition with saturation recovery. Magn Reson Med 2012;67(3):622–627.

10.

Guyader

J-M

Huizinga

Fortunati

, et al. Total correlation-based groupwise image registration for quantitative MRI. In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops, 2016, pp.186–193. IEEE (Institute of Electrical and Electronic Engineers).

11.

Zhang

Kabus

, et al. Cardiac magnetic resonance T1 and extracellular volume mapping with motion correction and co-registration based on fast elastic image registration. MAGMA 2018;31:115–129.

12.

Piechnik

Ferreira

Dall'Armellina

Cochlin

Greiser

Neubauer

, et al. Shortened Modified Look-Locker Inversion recovery (ShMOLLI) for clinical myocardial T1-mapping at 1.5 and 3 T within a 9 heartbeat breathhold. J Cardiovasc Magn Reson 2010;12(1):69.

13.

El-Rewaidy

Nezafat

Jang

, et al. Nonrigid active shape model–based registration framework for motion correction of cardiac T1 mapping. Magn Reson Med 2018;80:780–791.

14.

Tilborghs

Dresselaers

Claus

, et al. Robust motion correction for cardiac T1 and ECV mapping using a T1 relaxation model approach. Med Image Anal 2019;52:212–227.

15.

Dabir

Child

Kalra

, et al. Reference values for healthy human myocardium using a T1 mapping methodology: results from the international T1 multicenter cardiovascular magnetic resonance study. J Cardiovasc Magn Reson 2014;16:69.

16.

Lee

Liu

Nacif

, et al. Myocardial T1 and extracellular volume fraction mapping at 3 tesla. J Cardiovasc Magn Reson 2011;13:75.

17.

Messroghli

Greiser

Fröhlich

, et al. Optimization and validation of a fully-integrated pulse sequence for modified look-locker inversion-recovery (MOLLI) T1 mapping of the heart. J Magn Reson Imaging 2007;26:1081–1086.

18.

Song

Stainsby

, et al. Flexible cardiac T1 mapping using a modified Look–locker acquisition with saturation recovery. Magn Reson Med 2012;67:622–627.

19.

Kawel

Nacif

Zavodni

, et al. T1 mapping of the myocardium: intra-individual assessment of post-contrast T1 time evolution and extracellular volume fraction at 3T for Gd-DTPA and Gd-BOPTA. J Cardiovasc Magn Reson 2012;14:26.

20.

Neilan

Coelho-Filho

Shah

, et al. Myocardial extracellular volume fraction from T1 measurements in healthy volunteers and mice: relationship to aging and cardiac dimensions. JACC Cardiovasc Imaging 2013;6:672–683.

21.

Piechnik

Ferreira

Dall’Armellina

, et al. Shortened modified look-locker inversion recovery (ShMOLLI) for clinical myocardial T1-mapping at 1.5 and 3 T within a 9 heartbeat breathhold. J Cardiovasc Magn Reson 2010;12:69.

22.

von Knobelsdorff-Brenkenhoff

Prothmann

Dieringer

, et al. Myocardial T 1 and T 2 mapping at 3 T: reference values, influencing factors and implications. J Cardiovasc Magn Reson 2013;15:53.

23.

Kellman

Wilson

Xue

, et al. Extracellular volume fraction mapping in the myocardium, part 2: initial clinical experience. J Cardiovasc Magn Reson 2012;14:64.

24.

Kellman

Hansen

. T1-mapping in the heart: accuracy and precision. J Cardiovasc Magn Reson 2014;16:2.

25.

Ugander

Oki

Hsu

L-Y

, et al. Extracellular volume imaging by magnetic resonance imaging provides insights into overt and sub-clinical myocardial pathology. Eur Heart J 2012;33:1268–1278.

26.

Cerqueira

Weissman

Dilsizian

, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the cardiac imaging committee of the council on clinical cardiology of the American heart association. Circulation 2002;105:539–542.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.90 MB