Validation of a radial 3D cine sequence in a paediatric population

Background Recently, a 3D cine cardiovascular magnetic resonance (CMR) imaging sequence was developed to assess fetal cardiovascular malformations (1). Planning 2D slices in this population is particularly challenging due to the complex cardiac anatomy, high fetal heart rates, the small size of the heart relative to the large field of view, and impredictable fetal motion. The 3D cine CMR sequence overcomes these limitations by capturing the entire heart volume in a single acquisition, reducing planning time and enabling retrospective reconstruction of any desired 2D view. Validation is needed to incorporate this sequence in the clinical workflow. However, there is no golden standard for cardiac volumes in fetal CMR. We therefore validate in this study the 3D CMR sequence in a paediatric population with complex cardiac anatomy and relatively high heart rates. Methods The study was approved by the Regional Ethics Board. Informed consent was obtained from the participants' legal guardians. All examinations were conducted using a 1.5 T MR scanner in 8 participants (4 male), with ages 4-46 months. All subjects were either under general anaesthesia or sedated with dexmedetomidine. A 3D slab-selective radial balanced steady-state free precession (bSSFP) research pulse sequence was used for data acquisition in a paediatric population(1), with an isotropic voxel size of 1.6 mm3. For comparison, the standard clinical 2D Cartesian bSSFP cine sequence was acquired with typical voxel size: 0.8x0.8x5.0 mm³. The 3D cine volume was reconstructed and standardized cardiac views were obtained using multi-planar reformatting. End-diastolic volume and end-systolic volume were measured for the systemic ventricle in 3D cine images and compared to 2D cine images as reference standard. Agreement was assessed according to Bland-Altman with reported median bias and 95 % limits of agreement (2.5 – 97.5 percentiles). Results A 3D cine was successfully reconstructed for all 8 included participants with cardiac malformations. Typical acquisition times were 3.5 min for the 3D and around 10 minutes for obtaining the desired 2D views. Figure 1 shows a typical example comparing the two sequences. Note that signal intensities are lower for the 3D, which is inherent to the inflow effect (2). Figure 2 shows quantitative measurements. Conclusions The 3D CMR sequence demonstrates strong agreement with the clinical 2D standard for ventricular volume measurements, supporting its validity for volumetric assessment. These findings suggest that the 3D sequence may offer a reliable alternative for advanced cardiac imaging in clinical and research settings for shortening acquisition time, which in turn may decrease the need for sedation.