Fig 1.
Schematics of the TH-based exchange model.
CAIF, blood concentration of AIF; ha, arterial transfer function; Ca, concentration of the arterial plasma input to the tissue; Fp, plasma flow per unit tissue volume; Cp, concentration in the tissue capillary plasma space; Ce, concentration in the extravascular space; vp, plasma volume in the tissue; ve, extravascular interstitial volume; PS, product of the permeability and the surface area between the compartments; xi, a position along the capillary.
Table 1.
Definitions of perfusion parameters.
Fig 2.
Display of the 16 cardiac short-axis segments divided into epicardial and endocardial regions (total 32 segments).
Colors indicate the three coronary artery territories of the left anterior descending (LAD), right coronary artery (RCA), and the left circumflex coronary artery (LCX).
Table 2.
NTHf and AATH results: MBF and MPRI corresponding to percent diameter stenosis of coronary lesion.
Fig 3.
(a) Selective coronary angiography in a representative patient (83-year-old female) with angiographically normal coronary arteries. (b) First-pass perfusion mid-level MR image at stress. Bullseye plots of MBF (ml/g/min) (c) at rest and (d) at stress, and (e) MPRI calculated from the NTHf method.
Fig 4.
(a) Selective coronary angiography in a representative patient (66-year-old male) with severe triple-vessel disease (arrows). (b) First-pass enhancement mid-level MR image at stress showing a qualitatively delayed and decreased delivery of contrast agent to the inferior segments and milder delayed delivery to the anterior segments. Bullseye plots of MBF (ml/g/min) (c) at rest and (d) at stress, and (e) MPRI calculated from the NTHf method.
Fig 5.
(a) Selective coronary angiography in a representative patient (65-year-old male) with severe triple-vessel disease (arrows). (b) First-pass perfusion mid-level MR image at stress showing relatively decreased flow to the anterior and anteroseptal wall. Bullseye plots of MBF (ml/g/min) (c) at rest and (d) at stress, and (e) MPRI calculated from the NTHf method, showing more extensive perfusion abnormality.
Fig 6.
(a) Bullseye plot indicating representative segments of normal perfusion (blue) and perfusion defect (gray) in the patient with severe triple-vessel disease in Fig 5. Gd-DTPA concentration ([Gd-DTPA]) first-pass perfusion time-curves at (b) rest and (c) stress in the normal (blue circle) and defect (gray triangle) myocardium segments shown in (a) with best fit estimated plots from the NTHf (solid line) and AATH (dashed line) models. Estimated MBFs (ml/g/min) using the NTHf model at rest (stress) were 0.81 (1.56) in a normal segment and 0.99 (0.49) in a perfusion defect segment, and estimated MBFs using the AATH model at rest (stress) were 0.97 (1.30) in a normal segment and 0.93 (0.57) in a perfusion defect segment.
Fig 7.
The correlation plots of MBFs from the NTHf and AATH at (a) rest and (b) stress. (c) The correlation plot of MPRIs from the NTHf and AATH. p < 0.001 for all. Bland-Altman plots of MBFs from the NTHf and AATH at (a) rest and (b) stress. (c) Bland-Altman plot of MPRIs from the NTHf and AATH.
Fig 8.
Plots simulated with (top) MBF = 0.81 and (bottom) MBF = 2.2. (a) Simulated (or original; solid line), White Gaussian noise added (circle), and estimated from NTHf (dashed line) plots are shown at SNR = 10, 20, 30. (b) Percent errors for a range of SNR = 0–60 showing average and two standard deviation error bars (each SNR was repeated 30 times).