Table 1.
Patient demographics and clinical indications for cardiac MRI.
Fig 1.
Illustration of blood-pool to lung SNR and CNR calculations in a mid-ventricle slice.
The image is acquired using T1-TFE sequence. Two regions of interest (region 1 = blood pool, region 2 = lung) were manually chosen using circular region chosen tool in cvi42. Then the means, standard deviations as well as the medians of the selected regions will be reported, and these numbers are used for SNR and CNR calculations.
Fig 2.
Examples of images with different image quality scores for both bSSFP sequence and T1-TFE sequence.
In this study, all images receive scores no less than 2. So examples of grade 2, 3, and 4 are shown in this figure.
Fig 3.
Visual comparison of the images and the segmentations on a 17-year-old female patient with tetralogy of Fallot status post repair.
We showed the end-systolic and end-diastolic phases for the base slice, mid slice, and apex slice from both bSSFP sequence without contrast and T1-TFE sequence with Ferumoxytol.
Fig 4.
Statistical analysis for LVEDV, LVESV, LVSV, LVEF, LVEDV/BSA, RVEDV, RVESV, RVSV, RVEF, and RVEDV/BSA obtained from the bSSFP cine images and the T1-TFE cine images.
Bold lines indicate the bias, and dashed lines indicate the limits of agreement within a 95% confidence interval. Here, we plotted the Bland-Altman plots for these values. The statistical results showed that there is agreement and no significant difference between the cardiac function obtained using bSSFP and T1-TFE with contrast agent.
Table 2.
Mean ± SD of LV and RV dimensions based on acquisition with bSSFP and T1-TFE.
Table 3.
Statistical analysis of signal-to-noise and contrast-to-noise ratio.
Fig 5.
This figure presents a graphical representation of the subjective image assessment done by two readers.
Their respective p-values from paired t-test are included. The left side provides Reader 1’s ratings, and the right side provides Reader 2’s ratings.