Fig 1.
Light microscopy image of a histological preparation from the medial thigh muscle with vascular endothelium stained brown by isolectin.
The selection was performed by semiautomatic analysis of the images with the microscope’s morphometry software to determine the area, number and diameter of the vessels. The outer borders of the segmented vessels are labeled in green.
Fig 2.
Representative LMCA relaxation-rate-change time curves measured a) in the aorta and b) and c) in the medial thigh muscle are shown using black circles.
Fig a) shows the bolus-delay-corrected arterial input function (red solid line) related to the arterial hematocrit Hcta). Fig b) demonstrates the curve fitting result with the ETM and Fig c) the result with the 2CXM (red solid lines) and individual results for the plasma fraction (blue dot-dashed line) and the interstitial contribution (green dotted line). Obviously, the data are much better fitted using the 2CXM. Taking bolus broadening into account, the 2CXM yields a significantly larger area under the curve for the plasma contribution compared to the ETM, where the plasma compartment is merely the AIF scaled down with the plasma volume. As a result of the larger blood plasma fraction, the interstitial contribution using the 2CXM is lower, especially during the CA uptake period.
Fig 3.
BPCA data recorded from CA-induced time-dependent relaxation rate changes in a) the aorta (related to the arterial hematocrit level Hcta)) and b) the medial thigh muscle (black circles) of the same experiment as the examples in Fig 2.
Additionally, in a) the time delay and bolus-dispersion-corrected AIF and in b) one example of the fitting results with the BD method (red solid lines) is shown. As indicated by the statistical results (Table 1), vb is determined lower by BD compared with 2CXM (Fig 2) and EqMRI (Fig 4).
Fig 4.
For the EqMRI method, the same BPCA relaxation rate changes as in Fig 3 are shown as black circles.
No correction with respect to the arterial hematocrit level was performed at this point. Red solid lines represent the estimated average equilibrium states of a) the AIF in the aorta and in b) the medial thigh muscle (black circles) according to Eq (14) are shown.
Table 1.
Median values of the model parameter vb with first (Q1) and third quartiles (Q3) and interquartile ranges (IQR) obtained with the different models and methods in the different skeletal muscle regions of the pigs’ hind legs.
Also shown are the mean values of the contrast-to-noise ratio (CNR) of the CA time courses for each muscle region. The mean CNRs of the AIFs were 80 ± 40 (92 ± 37) for the BPCA (LMCA) measurements, respectively.
Fig 5.
Box plots of the model parameter vb obtained with the different models and methods in the different skeletal muscle regions of the pigs’ hind legs.
Table 2.
Spearman correlation coefficients of the individual results of the four MRI methods for the muscle groups in the hind legs of the pigs and Wicoxon rank sum test results between the different MRI methods and between MRI and histology.
Table 3.
Comparison of blood and plasma volume fractions of skeletal muscle of different species reported in the literature.