Comparison of a multi-breath-hold and a single breath-hold cine imaging approach for 4D guide-point modeling of the left ventricle

Purpose: Guide-point modeling (GPM) enables reliable and time-efficient assessment of left ventricular (LV) volumes when using sequences that allow acquisition of short- and long-axis scans within a single breath-hold. Slice misalignment may influence GPM analysis of standard multi-breath-hold images due to image acquisition in different breath-holds. Thus, our study aimed to assess if such an approach allows for reliable volumetric calculations in the clinical routine. Materials and Methods: 52 patients were examined on a 1.5 T scanner with multi-breath-hold acquisitions on the standard short- and long-axis using an SSFP (TR 3 ms, TE 1.5 ms, FA 60 ) sequence and a TPAT accelerated SSFP (TR 4.6 msec, TE 1.1msec, FA 60 , acceleration factor 3) sequence that covered the LV in 3 short- and 2 long-axis slices within a single breath-hold. For both datasets GPM was used to assess LV volumes. In addition, LV parameters were calculated by applying the summation of slices (SoS) approach (standard of reference) with the short-axis views of the multi-breath-hold dataset. Results: The post-processing times were shorter with both GPM approaches (both, p < 0.001). No significant difference between the 3 methods for the calculation of the ejection fraction was observed. However, end-diastolic, end-systolic and stroke volumes yielded higher results than the standard of reference if the GPM technique was employed. Excellent correlations were observed for all volumetric parameters derived from both GPM evaluations (all r > 0.97). Conclusion: Cine short- and long-axis images that had been acquired in different breath-holds can be reliably evaluated by the GPM approach. (orig.)