Limits...
Validation of Perfusion Quantification with 3D Gradient Echo Dynamic Contrast-Enhanced Magnetic Resonance Imaging Using a Blood Pool Contrast Agent in Skeletal Swine Muscle.

Hindel S, Sauerbrey A, Maaß M, Maderwald S, Schlamann M, Lüdemann L - PLoS ONE (2015)

Bottom Line: The F-test for model comparison was used to decide whether to use the results of the one- or two-compartment model fit.The resulting values of delay, dispersion, blood volume, mean transit time, and flow were all in physiologically and physically reasonable ranges.The DCE-MRI technique presented here uses a blood pool contrast medium in combination with a two-compartment tracer kinetic model and allows absolute quantification of low-perfused non-cerebral organs such as muscles.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiotherapy, Medical Physics, University Hospital Essen, Essen, North Rhine-Westphalia, Germany.

ABSTRACT
The purpose of our study was to validate perfusion quantification in a low-perfused tissue by dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) with shared k-space sampling using a blood pool contrast agent. Perfusion measurements were performed in a total of seven female pigs. An ultrasonic Doppler probe was attached to the right femoral artery to determine total flow in the hind leg musculature. The femoral artery was catheterized for continuous local administration of adenosine to increase blood flow up to four times the baseline level. Three different stable perfusion levels were induced. The MR protocol included a 3D gradient-echo sequence with a temporal resolution of approximately 1.5 seconds. Before each dynamic sequence, static MR images were acquired with flip angles of 5°, 10°, 20°, and 30°. Both static and dynamic images were used to generate relaxation rate and baseline magnetization maps with a flip angle method. 0.1 mL/kg body weight of blood pool contrast medium was injected via a central venous catheter at a flow rate of 5 mL/s. The right hind leg was segmented in 3D into medial, cranial, lateral, and pelvic thigh muscles, lower leg, bones, skin, and fat. The arterial input function (AIF) was measured in the aorta. Perfusion of the different anatomic regions was calculated using a one- and a two-compartment model with delay- and dispersion-corrected AIFs. The F-test for model comparison was used to decide whether to use the results of the one- or two-compartment model fit. Total flow was calculated by integrating volume-weighted perfusion values over the whole measured region. The resulting values of delay, dispersion, blood volume, mean transit time, and flow were all in physiologically and physically reasonable ranges. In 107 of 160 ROIs, the blood signal was separated, using a two-compartment model, into a capillary and an arteriolar signal contribution, decided by the F-test. Overall flow in hind leg muscles, as measured by the ultrasound probe, highly correlated with total flow determined by MRI, R = 0.89 and P = 10-7. Linear regression yielded a slope of 1.2 and a y-axis intercept of 259 mL/min. The mean total volume of the investigated muscle tissue corresponds to an offset perfusion of 4.7mL/(min ⋅ 100cm3). The DCE-MRI technique presented here uses a blood pool contrast medium in combination with a two-compartment tracer kinetic model and allows absolute quantification of low-perfused non-cerebral organs such as muscles.

No MeSH data available.


Related in: MedlinePlus

Regression of the corrected DCE-MRI models.Regression of the corrected DCE-MRI model results with the Doppler flow values for the one-compartment model (A) and for the F-test selection method (B). The solid line represents the linear regression fit of the measured data (circles). The dotted line depicts the standard deviation of the data related to the linear regression.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4465215&req=5

pone.0128060.g008: Regression of the corrected DCE-MRI models.Regression of the corrected DCE-MRI model results with the Doppler flow values for the one-compartment model (A) and for the F-test selection method (B). The solid line represents the linear regression fit of the measured data (circles). The dotted line depicts the standard deviation of the data related to the linear regression.

Mentions: We plotted 20 flow values from seven pigs measured with the DCE-MRI model versus the corresponding Doppler flow values (Fig 8). If only the one-compartment model is used for all segments (Fig 8A), a correlation of R = 0.51 is obtained with a significance of P = 0.02, a slope of the regression line of m = 0.4, and an ordinate axis intercept of the regression line of 371 mL/min. Ideally, a slope of 1 and an ordinate intercept of zero would be expected. If we take into account both the one-compartment and the two-compartment model and let the F-test decide which one delivers the most accurate results (Fig 8B), we obtain a correlation of the flow values of R = 0.89 with a significance of P = 10−7, a slope of the regression line of 1.2, and an ordinate intercept of 259 mL/min. The ordinate intercept of 259 mL/min represents an average perfusion of the hind leg of 4.7 mL/(min⋅100cm3).


Validation of Perfusion Quantification with 3D Gradient Echo Dynamic Contrast-Enhanced Magnetic Resonance Imaging Using a Blood Pool Contrast Agent in Skeletal Swine Muscle.

Hindel S, Sauerbrey A, Maaß M, Maderwald S, Schlamann M, Lüdemann L - PLoS ONE (2015)

Regression of the corrected DCE-MRI models.Regression of the corrected DCE-MRI model results with the Doppler flow values for the one-compartment model (A) and for the F-test selection method (B). The solid line represents the linear regression fit of the measured data (circles). The dotted line depicts the standard deviation of the data related to the linear regression.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4465215&req=5

pone.0128060.g008: Regression of the corrected DCE-MRI models.Regression of the corrected DCE-MRI model results with the Doppler flow values for the one-compartment model (A) and for the F-test selection method (B). The solid line represents the linear regression fit of the measured data (circles). The dotted line depicts the standard deviation of the data related to the linear regression.
Mentions: We plotted 20 flow values from seven pigs measured with the DCE-MRI model versus the corresponding Doppler flow values (Fig 8). If only the one-compartment model is used for all segments (Fig 8A), a correlation of R = 0.51 is obtained with a significance of P = 0.02, a slope of the regression line of m = 0.4, and an ordinate axis intercept of the regression line of 371 mL/min. Ideally, a slope of 1 and an ordinate intercept of zero would be expected. If we take into account both the one-compartment and the two-compartment model and let the F-test decide which one delivers the most accurate results (Fig 8B), we obtain a correlation of the flow values of R = 0.89 with a significance of P = 10−7, a slope of the regression line of 1.2, and an ordinate intercept of 259 mL/min. The ordinate intercept of 259 mL/min represents an average perfusion of the hind leg of 4.7 mL/(min⋅100cm3).

Bottom Line: The F-test for model comparison was used to decide whether to use the results of the one- or two-compartment model fit.The resulting values of delay, dispersion, blood volume, mean transit time, and flow were all in physiologically and physically reasonable ranges.The DCE-MRI technique presented here uses a blood pool contrast medium in combination with a two-compartment tracer kinetic model and allows absolute quantification of low-perfused non-cerebral organs such as muscles.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiotherapy, Medical Physics, University Hospital Essen, Essen, North Rhine-Westphalia, Germany.

ABSTRACT
The purpose of our study was to validate perfusion quantification in a low-perfused tissue by dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) with shared k-space sampling using a blood pool contrast agent. Perfusion measurements were performed in a total of seven female pigs. An ultrasonic Doppler probe was attached to the right femoral artery to determine total flow in the hind leg musculature. The femoral artery was catheterized for continuous local administration of adenosine to increase blood flow up to four times the baseline level. Three different stable perfusion levels were induced. The MR protocol included a 3D gradient-echo sequence with a temporal resolution of approximately 1.5 seconds. Before each dynamic sequence, static MR images were acquired with flip angles of 5°, 10°, 20°, and 30°. Both static and dynamic images were used to generate relaxation rate and baseline magnetization maps with a flip angle method. 0.1 mL/kg body weight of blood pool contrast medium was injected via a central venous catheter at a flow rate of 5 mL/s. The right hind leg was segmented in 3D into medial, cranial, lateral, and pelvic thigh muscles, lower leg, bones, skin, and fat. The arterial input function (AIF) was measured in the aorta. Perfusion of the different anatomic regions was calculated using a one- and a two-compartment model with delay- and dispersion-corrected AIFs. The F-test for model comparison was used to decide whether to use the results of the one- or two-compartment model fit. Total flow was calculated by integrating volume-weighted perfusion values over the whole measured region. The resulting values of delay, dispersion, blood volume, mean transit time, and flow were all in physiologically and physically reasonable ranges. In 107 of 160 ROIs, the blood signal was separated, using a two-compartment model, into a capillary and an arteriolar signal contribution, decided by the F-test. Overall flow in hind leg muscles, as measured by the ultrasound probe, highly correlated with total flow determined by MRI, R = 0.89 and P = 10-7. Linear regression yielded a slope of 1.2 and a y-axis intercept of 259 mL/min. The mean total volume of the investigated muscle tissue corresponds to an offset perfusion of 4.7mL/(min ⋅ 100cm3). The DCE-MRI technique presented here uses a blood pool contrast medium in combination with a two-compartment tracer kinetic model and allows absolute quantification of low-perfused non-cerebral organs such as muscles.

No MeSH data available.


Related in: MedlinePlus