Blood T1* correction increases accuracy of extracellular volume measurements using 3T cardiovascular magnetic resonance: Comparison of T1 and T1* maps

Automated quantification of myocardial tissue characteristics from native T1 mapping using neural networks with uncertainty-based quality-control

BACKGROUND

Tissue characterisation with cardiovascular magnetic resonance (CMR) parametric mapping has the potential to detect and quantify both focal and diffuse alterations in myocardial structure not assessable by late gadolinium enhancement. Native T1 mapping in particular has shown promise as a useful biomarker to support diagnostic, therapeutic and prognostic decision-making in ischaemic and non-ischaemic cardiomyopathies.

METHODS

Convolutional neural networks (CNNs) with Bayesian inference are a category of artificial neural networks which model the uncertainty of the network output. This study presents an automated framework for tissue characterisation from native shortened modified Look-Locker inversion recovery ShMOLLI T1 mapping at 1.5 T using a Probabilistic Hierarchical Segmentation (PHiSeg) network (PHCUMIS 119-127, 2019). In addition, we use the uncertainty information provided by the PHiSeg network in a novel automated quality control (QC) step to identify uncertain T1 values. The PHiSeg network and QC were validated against manual analysis on a cohort of the UK Biobank containing healthy subjects and chronic cardiomyopathy patients (N=100 for the PHiSeg network and N=700 for the QC). We used the proposed method to obtain reference T1 ranges for the left ventricular (LV) myocardium in healthy subjects as well as common clinical cardiac conditions.

RESULTS

T1 values computed from automatic and manual segmentations were highly correlated (r=0.97). Bland-Altman analysis showed good agreement between the automated and manual measurements. The average Dice metric was 0.84 for the LV myocardium. The sensitivity of detection of erroneous outputs was 91%. Finally, T1 values were automatically derived from 11,882 CMR exams from the UK Biobank. For the healthy cohort, the mean (SD) corrected T1 values were 926.61 (45.26), 934.39 (43.25) and 927.56 (50.36) for global, interventricular septum and free-wall respectively.

CONCLUSIONS

The proposed pipeline allows for automatic analysis of myocardial native T1 mapping and includes a QC process to detect potentially erroneous results. T1 reference values were presented for healthy subjects and common clinical cardiac conditions from the largest cohort to date using T1-mapping images.

Towards measuring the effect of flow in blood T 1 assessed in a flow phantom and in vivo

Measurement of the blood T ₁ time using conventional myocardial T ₁ mapping methods has gained clinical significance in the context of extracellular volume (ECV) mapping and synthetic hematocrit (Hct). However, its accuracy is potentially compromised by in-flow of non-inverted/non-saturated spins and in-flow of spins which are not partially saturated from previous imaging pulses. Bloch simulations were used to analyze various flow effects separately. T ₁ measurements of gadolinium doped water were performed using a flow phantom with adjustable flow velocities at 3 T. Additionally, in vivo blood T ₁ measurements were performed in 6 healthy subjects (26 ± 5 years, 2 female). To study the T ₁ time as a function of the instantaneous flow velocity, T ₁ times were evaluated in an axial imaging slice of the descending aorta. Velocity encoded cine measurements were performed to quantify the flow velocity throughout the cardiac cycle. Simulation results show more than 30% loss in accuracy for 10% non-prepared in-flowing spins. However, in- and out-flow to the imaging plane only demonstrated minor impact on the T ₁ time. Phantom T ₁ times were decreased by up to 200 ms in the flow phantom, due to in-flow of non-prepared spins. High flow velocities cause in-flow of spins that lack partial saturation from the imaging pulses but only lead to negligible T ₁ time deviation (less than 30 ms). In vivo measurements confirm a substantial variation of the T ₁ time depending on the flow velocity. The highest aortic T ₁ times are observed at the time point of minimal flow with increased flow velocity leading to reduction of the measured T ₁ time by up to [Formula: see text] at peak velocity. In this work we attempt to dissect the effects of flow on T ₁ times, by using simulations, well-controlled, simplified phantom setup and the linear flow pattern in the descending aorta in vivo.

Oxygenation-sensitive cardiovascular magnetic resonance in hypertensive heart disease with left ventricular myocardial hypertrophy and non-left ventricular myocardial hypertrophy: Insight from altered mechanics and cardiac BOLD imaging

BACKGROUND

BOLD (blood oxygen level dependent) MRI can detect regional condition of myocardial oxygen supply and demand by means of paramagnetic properties.

PURPOSE

Noninvasive assessment of myocardial oxygenation by BOLD MRI in hypertensive patients with hypertension (HTN) left ventricular myocardial hypertrophy (LVMH) and HTN non-LVMH and its correlation with myocardial mechanics were performed.

STUDY TYPE

Prospective.

POPULATION

Twenty patients with HTN LVMH, 21 patients with HTN non-LVMH, and 23 normotensive controls were enrolled.

FIELD STRENGTH/SEQUENCE

Cine imaging, T2* and T1 mapping sequences were achieved at 3.0T.

ASSESSMENT

Dedicated T1 mapping, T2*, and cine imaging analysis were performed by two radiologists using cvi42.

STATISTICAL TESTS

One-way analysis of variance, Kruskal-Wallis test, Bland-Altman analysis, Pearson's correlation coefficient, Spearman's rank correlation.

RESULTS

T2* values of HTN LVMH group were significantly lower versus the controls (23.78 ± 3.09 versus 30.77 ± 2.71; P < 0.001) and HTN non-LVMH group (23.78 ± 3.09 versus 28.64 ± 4.23; P < 0.001). Left ventricular peak circumferential strain were reduced in HTN LVMH patients compared with other two groups (-11.32 [-15.64, -10.3], -16.78 [-19.35, -15.34], and -19.73 [-20.57, -18.73]; P < 0.05); and longitudinal strain of HTN LVMH patients were lower than other two groups (-11.31 ± 2.91, -15.1 ± 3.06, and -18.85 ± 1.85; P < 0.05); radial strain of HTN LVMH patients were also lower than other two groups (25.03 ± 16, 40.95 ± 17.5 and 47.9 ± 10.23; P < 0.05). Extracellular volume correlated with peak circumferential, longitudinal, and radial strain (spearman rho = 0.6, 0.64, and -0.69; P < 0.05), respectively; T2* negatively correlated with peak circumferential and longitudinal strain (spearman rho = -0.43 and -0.49; P < 0.05), respectively. Patients with lower T2* values had significant decreases in myocardial mechanics (P < 0.05).

DATA CONCLUSION

HTN LVMH patients have both impaired myocardial mechanics and decreased T2* values compared with HTN non-LVMH and normotensive groups. BOLD MRI could provide a feasible assessment modality for detecting altered T2* due to the change of de-oxygenated hemoglobin and hence to the change of signal intensity in oxygenation-sensitive images.

LEVEL OF EVIDENCE

1 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2018;47:1297-1306.