Cardiac magnetic resonance imaging in small animal models of human heart failure.

The aim of this study was to test the feasibility of cine magnetic resonance imaging (MRI) for assessment of the infarcted rat and mouse heart and to compare the results with established methods. These models have been proven to predict genesis and prevention of heart failure in patients. The value of cine MRI was tested in studies investigating interventions to change the course of the remodeling process. MRI was performed for determination of left ventricular (LV) volumes and mass, myocardial infarct (MI) size and cardiac output. LV wet weight was determined after MRI. Rats underwent conventional hemodynamic measurements for determination of cardiac output and LV volumes by electromagnetic flowmeter and pressure-volume curves. Infarct size was determined by histology. MRI-acquired MI-size (18.5+/-2%) was smaller than that found by histology (22.8+/-2.5%, p<0.05) with close correlation (r=0.97). There was agreement in LV mass between MRI and wet weight (r=0.97, p<0.05) and in the MRI- and flowmeter measurements of cardiac output (r=0.80, p<0.05). Volume by MRI differed from pressure-volume curves with good correlation (r=0.96, p<0.05). In a serial study of mice after coronary ligation, LV hypertrophy at 8 weeks was detected (Sham 105.1+/-7.9 mg, MI 144.4+/-11.7 mg, p<0.05). Left ventricles were enlarged in infarcted mice (end-diastolic volume, week 8: Sham 63.5+/-4 microl, MI 94.2 microl, p<0.05). In conclusion, cine MRI is a valuable diagnostic tool applicable to the rat and mouse model of MI. Being non-invasive and exact it offers new insights into the remodeling process after MI because serial measurements are possible. The technique was applied to study several interventions and proved its usefulness.

Serial cine-magnetic resonance imaging of left ventricular remodeling after myocardial infarction in rats.

The purpose of the present study was the serial investigation of morphological and functional changes after left coronary artery ligation in the intact rat using cine-magnetic resonance imaging (MRI). MRI studies were performed 4, 8, 12, and 16 weeks after myocardial infarction (MI) with an echocardiogram (ECG)-triggered cine-fast low-angle shot (FLASH)-sequence in a 7-Tesla magnet. MI-size, left ventricular (LV) mass and volumes, cardiac index, ejection fraction (EF), and remote wall and scar thickness of 11 Wistar rats were compared to four sham-operated rats. Stress MRI with dobutamine (10 microl/kg x minute) was performed at 16 weeks. In MI groups (small MI < 30%, N = 5, large MI > 30%, N = 6), there was significant increase of LV mass (small MI + 47.8% increase, large MI + 74.1%) and wall thickness (large MI 1.21 +/- 0.03 to 1.84 +/- 0.07 mm). Scar thickness declined from four to 16 weeks (large MI 0.92 +/- 0.06 to 0.38 +/- 0.02 mm, P < 0.05). End-diastolic volume of both MI groups was significantly elevated but increased further only in animals with large MI from four to 16 weeks (657.1 +/- 38.6 to 869.7 +/- 60.7 microL, P < 0.05). Compared to sham, EF was significantly depressed in MI (large MI 31.5 +/- 2.0%). Wall thickening declined from four to 16 weeks post-MI (large MI 50.9 +/- 9.9 to 28.9 +/- 4.4%, P < 0.05). During stress, sham and MI rats increased wall thickening from 66.5 +/- 8.2 to 111.2 +/- 6.7% and from 30.8 +/- 4.3 to 47.5 +/- 5.8%, respectively (P < 0.05). Hypertrophy was found in all animals with MI throughout the entire period of observation, whereas dilatation after four weeks was only detected in animals with large MI. These morphologic changes were accompanied by an early decline of EF; myocardial function characterized by wall thickening deteriorated later.

Combined high-speed NMR imaging of perfusion and microscopic coronary conductance vessels in the isolated rat heart.

Noninvasive characterization of microcirculation at the level of both coronary conductance and resistance vessels is of major importance for the understanding of microvascular adaptive processes in the heart. The objective of this study was to determine simultaneously myocardial perfusion and microvessel diameters in the myocardium by magnetic resonance (MR) imaging within the same heart. A MR imaging method is presented which combines high-resolution perfusion measurement (140 x 140 microm2) by spin labeling with flow-weighted MR microscopy of coronary microvessels (phi > 140 microm). We determined changes in myocardial perfusion and vessel diameters of isolated beating rat hearts (n = 10) at rest and during administration of nitroglycerin (0.5 mg/min). Alterations in perfusion were validated by microsphere measurements. Under the influence of nitroglycerin an increase in perfusion (+2.51 +/- 0.4 ml x min(-1) x g(-1), mean +/- SEM) and vessel diameters (+14.22 +/- 1.92%) could be observed. Endocardial perfusion revealed a modest enhanced susceptibility to nitroglycerin in comparison to epicardial perfusion. Analysis of vessels according to their diameters showed no significant differences. MR imaging allows the noninvasive and simultaneous determination of conducting arteries and smaller resistance vessels in one and the same beating rat heart. Due to an excellent spatial resolution of these methods, transmural characterization of both parameters at rest and during vasodilation is feasible.

Myocardial perfusion imaging using a non-contrast agent MR imaging technique.

INTRODUCTION: A MR imaging (MRI) method has been developed to determine quantitatively myocardial perfusion (P) in the rat heart in vivo. This method has the potential to non-invasively measure cardiac perfusion without the use of a contrast agent by exploiting the endogenous contrast from flowing blood itself. METHOD AND RESULTS: Principle of the technique is the arterial spin labeling of endogenous water protons within the short axis imaging slice. Arterial spin labeling techniques are based on a model that uses inflow effects to relate intrinsic changes in longitudinal relaxation (T1) to tissue perfusion. Perfusion is determined from the difference between a slice selective and a global inversion recovery experiment. Perfusion was determined at rest and during hyperemia induced by intravenous adenosine (3 mg/(kg min)). The MR perfusion values were compared with perfusion data obtained in the same animal using the colored microspheres (MS) technique as the gold standard. The MR perfusion (mean +/- SEM) was 3.3 +/- 0.2 ml/min/g at rest and 4.6 +/- 0.6 ml/min/g during adenosine. Perfusion values obtained by colored MS were 3.4 +/- 0.2 and 4.7 +/- 0.8 ml/min/g at rest and during vasodilation, respectively. Adenosine decreased mean arterial pressure (MAP) from 120 to 65 mmHg which implies a reduction of coronary resistance (CR) to about 50% of baseline. CONCLUSION: Our study shows that quantitative mapping of perfusion may be performed non-invasively by MRI. The MR perfusion data are in excellent correlation with data obtained by the well-established colored MS technique. Determination of perfusion reserve confirms that coronary perfusion is highly dependent on blood pressure due to changes in CR.

Dobutamine-stress magnetic resonance microimaging in mice : acute changes of cardiac geometry and function in normal and failing murine hearts.

The aim of this study was to assess the capability of MRI to characterize systolic and diastolic function in normal and chronically failing mouse hearts in vivo at rest and during inotropic stimulation. Applying an ECG-gated FLASH-cine sequence, MRI at 7 T was performed at rest and after administration of 1.5 microgram/g IP dobutamine. There was a significant increase of heart rate, cardiac output, and ejection fraction and significant decrease of end-diastolic and end-systolic left ventricular (LV) volumes (P<0.01 each) in normal mice during inotropic stimulation. In mice with heart failure due to chronic myocardial infarction (MI), MRI at rest revealed gross LV dilatation. There was a significant decrease of LV ejection fraction in infarcted mice (29%) versus sham mice (58%). Mice with MI showed a significantly reduced maximum LV ejection rate (P<0.001) and LV filling rate (P<0.01) and no increase of LV dynamics during dobutamine action, indicating loss of contractile and relaxation reserve. In 4-month-old transgenic mice with cardiospecific overexpression of the beta(1)-adrenergic receptor, which at this early stage do not show abnormalities of resting cardiac function, LV filling rate failed to increase after dobutamine stress (transgenic, 0.19+/-0.03 microL/ms; wild type, 0.36+/-0.01 microL/ms; P<0.01). Thus, MRI unmasked diastolic dysfunction during dobutamine stress. Dobutamine-stress MRI allows noninvasive assessment of systolic and diastolic components of heart failure. This study shows that MRI can demonstrate loss of inotropic and lusitropic response in mice with MI and can unmask diastolic dysfunction as an early sign of cardiac dysfunction in a transgenic mouse model of heart failure.

Serial magnetic resonance imaging of microvascular remodeling in the infarcted rat heart.

BACKGROUND: Alterations in the coronary circulation are important determinants of myocardial function. Few data are available, however, about microvascular changes in reactive hypertrophy. With MRI, serial determination of myocardial microcirculation after myocardial infarction (MI) is feasible. METHODS AND RESULTS: We quantitatively determined myocardial perfusion and relative intracapillary blood volume using an MRI technique. Infarct size, myocardial mass, and left ventricular volumes were determined with cine MRI. Rats were investigated at 8, 12, and 16 weeks after MI (mean MI size 24.1+/-2.0%) or sham operation. Vasodilation was induced by adenosine. In the infarcted group, maximum perfusion decreased significantly from 8 to 16 weeks (5.6+/-0.3 versus 3.5+/-0.2 mL. g(-1). min(-1), P<0.01) compared with sham animals (5.5+/-0.3 versus 5.0+/-0.2 mL. g(-1). min(-1), P=0.17). Myocardial mass increased significantly (559.1+/-20.8 mg at 8 weeks versus 690.9+/-42.7 mg at 16 weeks, P<0.05) compared with sham-operated animals (516.3+/-41.7 versus 549.2+/-32.3 mg). Basal relative intracapillary blood volume increased significantly to 15.7+/-0.5 vol% at 8 weeks after MI and remained elevated (16.8+/-0.6 vol%) at 16 weeks compared with 12.1+/-0.3 vol% (P<0.01) in sham-operated rats. CONCLUSIONS: Our results indicate that significant microvascular changes occur during cardiac remodeling. Hypoperfusion in the hypertrophied myocardium is related to an increase in vascular capacity, suggesting a compensatory vasodilatory response at the capillary level. These microvascular changes may therefore contribute to the development of heart failure.

Fast high-resolution magnetic resonance imaging demonstrates fractality of myocardial perfusion in microscopic dimensions.

The fractal nature of heterogeneity of myocardial blood flow and its implications for the healthy and diseased heart is not yet understood. The main hindrance for investigation of blood flow heterogeneity and its role in physiology and pathophysiology is that conventional methods for determination of myocardial perfusion have severe limitations concerning temporal and spatial resolution and invasiveness. In isolated rat hearts, we developed a nuclear magnetic resonance technique that does not depend on contrast agents and in which the apparent longitudinal relaxation time is made perfusion sensitive by selective preparation of the imaging slice. This perfusion-sensitive relaxation time is determined within 40 seconds as a map with a high spatial in-plane resolution of 140x140 microm(2) and a thickness of 1.5 mm. Perfusion imaging was validated with the established microsphere technique. Additionally, the congruence between perfusion-sensitive T:(1) maps and first-pass perfusion imaging was demonstrated. As an application of high-resolution perfusion imaging, fractal analysis of the spatial distribution of perfusion was performed. We were able to demonstrate that the fractality of this distribution exists even in microscopic dimensions. Vasodilation by nitroglycerin modulated the fractal pattern of perfusion, and the decrease of the fractal dimension indicated a shift toward homogeneity. This implies that parameters of the fractal distribution depend on the microvascular tone rather than on anatomic preformations; ie, fractality is a functional characteristic of perfusion.

Developmental changes of cardiac function and mass assessed with MRI in neonatal, juvenile, and adult mice.

Cardiovascular transgenic mouse models with an early phenotype or even premature death require noninvasive imaging methods that allow for accurate visualization of cardiac morphology and function. Thus the purpose of our study was to assess the feasibility of magnetic resonance imaging (MRI) to characterize cardiac function and mass in newborn, juvenile, and adult mice. Forty-five C57bl/6 mice from seven age groups (3 days to 4 mo after birth) were studied by MRI under isoflurane anesthesia. Electrocardiogram-gated cine MRI was performed with an in-plane resolution of (78-117 microm)(2). Temporal resolution per cine frame was 8.6 ms. MRI revealed cardiac anatomy in mice from all age groups with high temporal and spatial resolution. There was close correlation between MRI- and autopsy-determined left ventricular (LV) mass (r = 0.95, SE of estimate = 9.5 mg). The increase of LV mass (range 9.6-101.3 mg), cardiac output (range 1.1-14.3 ml/min), and stroke volume (range 3. 2-40.2 microl) with age could be quantified by MRI measurements. Ejection fraction and cardiac index did not change with aging. However, LV mass index decreased with increasing age (P < 0.01). High-resolution MRI allows for accurate in vivo assessment of cardiac function in neonatal, juvenile, and adult mice. This method should be useful when applied in transgenic mouse models.

In vivo assessment of cardiac remodeling after myocardial infarction in rats by cine-magnetic resonance imaging.

The rat infarct model offers important parallels to the process of remodeling after myocardial infarction (MI) in humans. The aim of this study was to test the feasibility of cine fast low-angle shot (FLASH) magnetic resonance imaging (MRI) for assessment of the infarcted and noninfarcted rat heart and to compare the results with established methods. In group A, MRI was done 8-16 weeks after MI on a 7-T scanner using an electrocardiogram-triggered cine-FLASH sequence. We determined left ventricular (LV) volumes and mass, wall thickness, MI size, cardiac output, and ejection fraction. Afterward, MI size was histologically determined. In group B, after MRI eight controls and eight rats 16 weeks after MI underwent conventional hemodynamic measurements for determination of cardiac output, LV volumes, and ejection fraction by electromagnetic flowmeter and pressure-volume curves. LV wet weight was determined. In group A, MRI-acquired MI size (18.5 +/- 2%) was smaller than histology (22.8 +/- 2.5%, p < 0.05) with close correlation (r = 0.97). In group B, agreement in LV mass was found between MRI and wet weight (controls, 537.6 +/- 19.6 vs. 540.3 +/- 18.4 mg; MI, 865.1 +/- 39.2 vs. 865.1 +/- 41.3 mg; for the difference p = ns, r = 0.97, p < 0.05) and in the MRI and flowmeter measurements (cardiac output, controls 73.1 +/- 2.9 vs. 75.2 +/- 2.6 ml/min; MI 82.4 +/- 5.2 vs. 81.9 +/- 3.7 ml/min; for the difference p = ns, r = 0.80, p < 0.05). End-diastolic volume by MRI differed from pressure-volume curves with good correlation (controls, 343.9 +/- 8.4 vs. 262.7 +/- 12.8 microl; MI, 737.0 +/- 70.5 vs. 671.1 +/- 64.1 microl; p < 0.05 each, r = 0.96, p < 0.05). Cine-FLASH-MRI is a valuable diagnostic tool applicable to the rat model of MI. Being noninvasive and exact, it offers new insights in the remodeling process after MI because serial measurements are possible.

Myocardial perfusion and intracapillary blood volume in rats at rest and with coronary dilatation: MR imaging in vivo with use of a spin-labeling technique.

PURPOSE: To validate a magnetic resonance (MR) imaging technique that is not first pass and that reveals perfusion and regional blood volume (RBV) in the intact rat. MATERIALS AND METHODS: Measurement of perfusion was based on the perfusion-sensitive T1 relaxation after magnetic spin labeling of water protons. RBV was determined from steady-state measurements of T1 before and after administration of an intravascular contrast agent. The colored microsphere technique was used as a reference method for perfusion measurement. RBV and perfusion maps were obtained with the rats at rest and during administration of 3 mg of adenosine phosphate per kilogram of body weight per minute. RESULTS: At MR imaging, perfusion during resting conditions was 3.5 mL/g/min +/- 0.1 (SEM), and RBV was 11.6% +/- 0.6 (SEM). Adenosine phosphate significantly increased perfusion to 4.5 mL/g/min +/- 0.3 (SEM) and decreased mean arterial pressure from 120 mm Hg to 65 mm Hg, which implies a reduction of coronary resistance to 40% of baseline. RBV increased consistently to 23.8% +/- 0.6 (SEM). CONCLUSION: The study results show that quantitative mapping of perfusion and RBV may be performed noninvasively by means of MR imaging in the intact animal. The presented method allows determination of vasodilative and perfusion reserve, which reflects the in vivo regulation of coronary microcirculation for a given stimulus.

Perfusion-corrected mapping of cardiac regional blood volume in rats in vivo.

Measurement of regional blood volume (RBV) in the myocardium in vivo is important for the assessment of tissue viability and function. The method in this work is based on the acquisition of a T(1) map before and after intravascular contrast agent application. It is known that this method is influenced by perfusion that causes an overestimation of RBV values. In order to solve this problem, the new method is proposed which acquires T(1) maps with slice selective inversion pulses. Due to blood flow nonexcited spins enter the detection slice, which leads to an acceleration of the relaxation time. A model that divides tissue into two compartments is adapted to slice selective inversion in order to derive a simple expression for perfusion-corrected RBV. The aim of the study is to demonstrate the feasibility and accuracy of this technique for quantification of RBV in rat myocardium in vivo. RBV maps were obtained for five rats, and the reproducibility was determined by repeating the experiment several times. A mean RBV value of 12.8 +/- 0.7% (v/v) over all animals was obtained in the myocardium. The results were compared with RBV maps obtained with perfusion-sensitive RBV imaging in the same five rats and with first-pass RBV studies. In order to demonstrate the strength of the new method the vasodilator adenosine was administered and alterations in microcirculation were imaged. Magn Reson Med 42:500-506, 1999.

In vivo quantitative mapping of cardiac perfusion in rats using a noninvasive MR spin-labeling method.

Measurement of myocardial perfusion is important for the functional assessment of heart in vivo. Our approach is based on the modification of the longitudinal relaxation time T1 induced by magnetic spin labeling of endogenous water protons. Labeling is performed by selectively inverting the magnetization within the detection slice, and longitudinal relaxation is measured using a fast gradient echo MRI technique. As a result of blood flow, nonexcited spins enter the detection slice, which leads to an acceleration of the relaxation rate. Incorporating this phenomenon in a mathematical model that describes tissue as two compartments yields a simple expression that allows the quantification of perfusion from a slice-selective and a global inversion recovery experiment. This model takes into account the difference between T1 in blood and T1 in tissue. Our purpose was to evaluate the feasibility and reproducibility of this technique to map quantitatively myocardial perfusion in vivo in rats. Quantitative maps of myocardial blood flow were obtained from nine rats, and the reproducibility of the technique was evaluated by repeating the whole perfusion experiment four times. Evaluation of regions of interest within the myocardium yielded a mean perfusion value of 3.6 +/- .5 ml x min(-1) x g(-1) over all animals, which is in good agreement with previously reported literature values.

Quantitative regional blood volume studies in rat myocardium in vivo.

Many pathophysiological processes in the myocardium are in close relation to changes of the regional blood volume and regional myocardial blood flow or perfusion. Only few methods exist to obtain quantitative values for these parameters. Quantitative regional blood volume (RBV) studies in rat myocardium are presented using snapshot fast low angle shot (FLASH) inversion recovery T1 measurements with two different blood pool contrast agents, gadolinium diethylenetriaminopentaacetic acid (Gd-DTPA) albumin and Gd-DTPA polylysine. In contrast to previous attempts, each snapshot FLASH image acquisition was ECG-triggered under breathhold conditions. To measure relaxation times shorter than a heart cycle, each T1 sequence was repeated two times with different delays between inversion pulse and first image acquisition. The experiments were performed on a Bruker Biospec 70/21 using a homogeneous transmitter coil and a circularly polarized surface receiver coil, a special ECG trigger unit, and a respirator that is controlled by the pulse program. Based on a fast exchange model RBVm maps were calculated from the relaxation time maps for different concentrations of the two blood pool contrast agents. A significant dependence of the RBVm values on blood T1 was found. This is in accordance with a model that has been developed recently relating the dependence of RBVm on T1 of blood to perfusion. For Gd-DTPA albumin, the application of the model to the experimental data yields realistic values for RBV and perfusion. The values, which are in accordance with literature data, were obtained at highest contrast agent concentrations i.e., lowest relaxation times of blood (ca. 200 ms).

Magnetic resonance microimaging for noninvasive quantification of myocardial function and mass in the mouse.

The purpose of this work was to develop high-resolution cardiac magnetic resonance imaging techniques for the in vivo mouse model for quantification of myocardial function and mass. Eight male mice were investigated on a 7-Tesla MRI scanner. High-quality images in multiple short axis slices (in-plane resolution 117 microm2, slice thickness 1 mm) were acquired with an ECG-gated cine sequence. Left ventricular end-diastolic and end-systolic volumes and mass were calculated from segmented slice volumes. There was precise agreement of left ventricular mass determined ex vivo and by MRI. Intraobserver (5%) and interobserver (5%) variability of in vivo MR measurements were low.

The effect of perfusion on T1 after slice-selective spin inversion in the isolated cardioplegic rat heart: measurement of a lower bound of intracapillary-extravascular water proton exchange rate.

Many NMR measurements of cardiac microcirculation (perfusion, intramyocardial blood volume) depend on some kind of assumption of intracapillary-extravascular water exchange rate, e.g., fast exchange. The magnitude of this water exchange rate, however, is still unknown. The intention of this study was to determine a lower limit for this exchange rate by investigating the effect of perfusion on relaxation time. Studies were performed in the isolated perfused cardioplegic rat heart. After slice-selective inversion, the spin lattice relaxation rate of myocardium within the slice was studied as a function of perfusion and compared with a mathematical model which predicts relaxation rate as a function of perfusion and intracapillary-extravascular exchange rate. A linear relationship was found between relaxation rate T(-1) and perfusion P normalized by perfusate/tissue partition coefficient of water, lambda: deltaT(-1) = m x deltaP/lambda with 0.82 < or = m < or = 1.06. Insertion of experimental data in the model revealed that a lower bound of the exchange rate from intra- to extravascular space is 6.6 s(-1) (4.5 s(-1), P < 0.05), i.e., the intracapillary lifetime of a water molecule is less than 150 ms (222 ms, P < 0.05). Based on this finding, the T1 mapping after slice-selective inversion could become a valuable noncontrast NMR method to measure variations of perfusion.

Three-dimensional coronary angiography of the perfused rat heart.

The purpose of this work was to visualize the whole three-dimensional coronary artery tree of the perfused beating rat heart using three-dimensional MRI. The spatial resolution amounts to 140 microns. Also, vessels having smaller diameters could be detected. Different strategies for the visualization of the three-dimensional coronary angiograms including maximum intensity projection, data thresholding, and segmentation, were shown. The coronary artery tree was best visualized by hysteresis threshold segmentation and subsequent surface reconstruction.

Study of microcirculation by coloured microspheres and NMR-microscopy in isolated rat heart: effect of ischaemia, endothelin-1 and endothelin-1 antagonist BQ 610.

Although the investigation of coronary microcirculation is of great importance, available methods have severe restrictions. They do not allow the study of vasodynamics of resistance vessels and microscopic conductance vessels simultaneously in the isolated beating rat heart. We now demonstrate that the combined measurement of perfusion which reflects the state of resistance vessels and cross-sections of microscopic conductance vessels is feasible in the model of the isolated constant flow perfused rat heart. Perfusion measurement was based on injection of coloured microspheres. Cross-sections of microscopic conductance vessels (diameter >140 micron) were determined by NMR-microscopy by flow weighted imaging. Both methods were established recently by our group. The combined measurement was applied to hearts which were subjected to ischaemia and reperfusion (group 1: n=5, 15 min ischaemia/group 2: n=7, 30 min ischaemia/measurements before ischaemia and 15/30 min after reperfusion), 200 pmol endothelin-1 bolus application (group 3: n=6/measurements before and 5 min after drug application), continuous infusion of the endothelin-1 antagonist BQ 610 (group 4: n=6/measurements before and 20 min after onset of infusion), and 200 pmol endothelin-1 application superimposed on 20 min of continuous BQ 610 infusion (group 5: n=7/combined measurement before BQ 610 infusion and 5 min after endothelin-1 application). In group 1, 15 min reperfusion restored the pre-ischaemic perfusion state, whereas conductance vessels were dilated (80.8+/-2.6%), after 30 min reperfusion pre-ischaemic conditions were also restored for conductance vessels. In group 2, a redistribution of perfusion from left ventricular endocardium to the right ventricular wall was observed. Post-ischaemic rhythm disturbances made NMR-imaging in this group impossible. In group 3, a shift of perfusion from the left ventricular myocardium to the right ventricular wall was observed. Similarly, the cross-section of left ventricular conductance vessels decreased (-32.6+/-2.1%), whereas size of right ventricular vessels increased. In group 4, BQ 610 had no effect on perfusion nor on vessel size and antagonized the effect of endothelin-1 on perfusion and vessel size in group 5.

In vivo colored microspheres in the isolated rat heart for use in NMR.

Myocardial perfusion measurement with colored microspheres may become an alternative for radioactive microsphere techniques. We use and validate a spectrophotometric method that has been previously established for large animals in the isolated perfused rat heart. The perfusion system was adapted for use in a NMR microscope. Hearts were perfused with constant coronary flow that was adjusted to a coronary perfusion pressure of 100 mmHg. Homogeneous coronary inflow of microspheres was represented by equal distribution of microspheres of two different colors after simultaneous injection. Mean regional myocardial blood flow was 17.76 +/- 5.01 ml/min/g, mean wet heart weight was 1.13 +/- 0.34 g and mean global flow was 20.06 +/- 0.60 ml/min. Heart rate was 296 +/- 8.9 beats/min and left ventricular pressure was similar 5 min before (149.1 +/- 14.27 mmHg) and after (147.1 +/- 13.49 mmHg) microsphere injection. Microspheres of four colors that were injected sequentially, at various coronary flows, demonstrated linearity and reproducibility of the technique. A cumulative use of less than 90 000 microspheres showed no effect on hemodynamics especially on left ventricular pressure.