Axial slice

Anesthesia: Mice are anesthetized within the magnet using 1.5% isofluran in air (see Hardware) applied at a rate of 75 ml/min using a home-build nose cone. Injection narcosis (e.g. Hypnorm/Diazepam) is disadvantageous due to insufficient controllability as well as respiratory and cardio-depressive side effects of these substances. In contrast, using inhalation anesthesia as per description, physiological heart and respiration rates of ~600 min^-1 and ~100 min^-1, respectively, can be maintained for more than 1 hour. This is important to obtain reliable information on murine heart function in vivo.

Temperature control: Furthermore, it is important to maintain the body temperature of the mouse (37 °C) during the whole measurement. Of course, no heating coils should be brought into the magnet, so that an alternative way is needed to control temperature. Therefore, we set the magnet gradient cooling system (coupled to a water recirculator Haake UWK 45) to body temperature which is unproblematic, unless the pulse sequence used does not extremely heat up the gradient system. For these cases, we developed a water cooling system with a special profile flexible tube that ensures a very high thermal transfer. The thereby produced water signal can be eliminated by a saturation module integrated into the pulse sequence. A thermocouple (Pt10Rh-Pt) is used to rectally monitor the animal's body temperature.

ECG and respiration gating: Gating is realized by a trigger module (SA Instruments M 1025) directly connected to the spectrometer. To obtain images at defined time points within the heart cycle, triggering of the image sequence (FLASH, Fast Low Angle Shot) to an ECG signal is necessary. This signal can be derived by fixing fore- and hindpaws to small pediatric electrodes. Furthermore, a pressure sensor is used to avoid respiratory motion artifacts by synchronization of the expiratory signal plateau with the pulse sequence. Only QRS complexes (triggering to end-diastolic phase) within this plateau are used to initialize a data acquisition cycle. During a complete heart cycle we usually acquire 15 to 20 images resulting in a temporal resolution of 5-7 ms (assuming a heart rate of 600 min^-1). A field of view of 3x3 cm² and a matrix size of 256x256 yield a pixel size of 117x117 µm (at a slice thickness of 1 mm). Temporal and spatial resolution are completely sufficient to routinely determine functional heart parameters (see movies on the left).

Orthostasis: To exclude that the vertical position of the mouse during the measurement affects the physiological parameters, we monitored hemodynamic and contractile parameters of isofluran-anestethized mice for more than 1 hour (corresponding to the total acquisition time for heart imaging). These data provided by a 1.4 F Millar catheter showed that the vertical position of the mouse has no significant effect on the results.

Coronal Slice

While analysis of the MR images (see next section) provides important parameters for the assessment of heart function, we are additionally able to get in vivo information about heart energetics without changing the experimental setup. Spatial resolved ³¹P-NMR spectroscopy is used to assess the energetic situation in the mouse heart by means of phosphocreatine and ATP signal intensities. A more detailed description of this method can be found under "Energetics" in the section In vivo 2D mapping of the energy state.

As can be seen from the figure on the right, both endo- and epicardial areas are approximated by circles, and the difference between the two radii provides an averaged wall thickness for the midventricular slice. Here, the papillary muscles have to be included into the endocardial area, whereas they have to be excluded for LV volume determination, since they are assigned to the myocardium and therefore not to the endocardial area.

Typical values for mouse hearts are given in the last column. It has to be mentioned that these were obtained from C57BL/6 wild type mice and that they can considerably vary for different strains.

In cooperation with the group of Prof. Dr. H. Drexler (MHH Hannover, Cardiology and Angiology) we consecutively investigated mice with mycardial infarction which was induced by occlusion of the LAD (left anterior descending coronary artery) under open-chest conditions and isoflurane inhalation anesthesia. Transverse slice images of a mouse heart before (A) and 1 (B) , 2 (C), and 4 (D) weeks after ligation of the LAD show that the progression of the myocardial infarct can easily be monitored by successive MR measurements. Furthermore, changes in the dynamics of LV contraction can directly be assessed using the cine sequence that is acquired for each heart slice.

As can be seen the initially normal heart dilates from week to week after induction of the myocardial infarct. Furthermore, the ventricular wall obviously becomes very thin in the region of the ligated LAD (right bottom). Watching the movie of the cine sequence, it is clearly seen that during systole, almost no contraction of the myocardium can be observed. This leads to a strongly reduced ejection fraction of less than 20 %. Surprisingly, despite these limitations in heart function, the animals are moving (climbing!) within their cages without any obvious handicap.