Accelerating global left-ventricular function assessment in mice using reduced slice acquisition and three-dimensional guide-point modelling.

BACKGROUND: To investigate the utility of three-dimensional guide-point modeling (GPM) to reduce the time required for CMR evaluation of global cardiac function in mice, by reducing the number of image slices required for accurate quantification of left-ventricular (LV) mass and volumes. METHODS: Five female C57Bl/6 mice 8 weeks post myocardial infarction induced by permanent occlusion of the left coronary artery, and six male control (un-operated) C57Bl/6 mice, were subject to CMR examination under isoflurane anaesthesia. Contiguous short axis (SAX) slices (1 mm thick 7-9 slices) were obtained together with two long axis (LAX) slices in two chamber and four chamber orientations. Using a mathematical model of the heart to interpolate information between the available slices, GPM LV mass and volumes were determined using full slice (all SAX and two LAX), six slice (four SAX and two LAX) and four slice (two SAX and two LAX) analysis protocols. All results were compared with standard manual volumetric analysis using all SAX slices. RESULTS: Infarct size was 39.1±5.1% of LV myocardium. No significant differences were found in left ventricular mass and volumes between the standard and GPM full and six slice protocols in infarcted mice (113±10, 116±11, and 117±11 mg respectively for mass), or between the standard and GPM full, six and four slice protocols in control mice, (105±14, 106±10, 104±12, and 105±7 mg respectively for mass). Significant differences were found in LV mass (135±18 mg) and EF using the GPM four slice protocol in infarcted mice (p<0.05). CONCLUSION: GPM enables accurate analysis of LV function in mice with relatively large infarcts using a reduced six slice acquisition protocol, and in mice with normal/symmetrical left-ventricular topology using a four slice protocol.

Ultra-fast and accurate assessment of cardiac function in rats using accelerated MRI at 9.4 Tesla.

MRI can accurately and reproducibly assess cardiac function in rodents but requires relatively long imaging times. Therefore, parallel imaging techniques using a 4-element RF-coil array and MR sequences for cardiac MRI in rats were implemented at ultra-high magnetic fields (9.4 Tesla [T]). The hypothesis that these developments would result in a major reduction in imaging time without loss of accuracy was tested on female Wistar rats under isoflurane anesthesia. High-resolution, contiguous short-axis slices (thickness 1.5 mm) were acquired covering the entire heart. Two interleaved data sets (i) with the volume coil (eight averages) and (ii) with the four-element coil array (one average) were obtained. In addition, two-, three-, and fourfold accelerated data sets were generated through postprocessing of the coil array data, followed by a TGRAPPA reconstruction, resulting in five data sets per rat (in-plane voxel size 100 x 100 microm). Using a single blinded operator, excellent agreement was obtained between volume coil (acquisition time: 88 min) and the fourfold accelerated (<3 min) data sets (e.g., LV mass 436 +/- 21 mg vs 433 +/- 19 mg; ejection fraction 74 +/- 5% vs 75 +/- 4%). This finding demonstrates that it is possible to complete a rat cine-MRI study under 3 min with low variability and without losing temporal or spatial resolution, making high throughput screening programs feasible.