3D models are accurate replicas of the cardiovascular anatomy and improve the understanding of complex CHD. 3D models did not change the surgical decision in most of the cases (21 of 40 cases, 52.5% cases). However, in 19 of the 40 selected complex cases, 3D model helped redefining the surgical approach.

Two-dimensional (2D) breath-hold cine MRI is used to assess cardiac anatomy and function. However, this technique requires cooperation from the patient, and in some cases the scan planning is complicated. Isotropic nonangulated threedimensional (3D) cardiac MR can overcome some of these problems because it requires minimal planning and can be reformatted in any plane. However, current methods, even those that use undersampling techniques, involve breath-holding for periods that are too long for many patients. Free-breathing respiratory gating sequences represent a possible solution for realizing 3D cine imaging. A real-time respiratory self-gating technique for whole-heart cine MRI is presented. Key words: whole-heart imaging; respiratory gating; cine MRI; free breathing; self navigation Two-dimensional (2D) cine imaging has been shown to be an accurate method of assessing cardiac anatomy and function (1). Unfortunately, it requires multiple breathholds and the scan planning requires operator knowledge of cardiac anatomy. Isotropic nonangulated three dimensional (3D) cardiac MR can overcome some of these problems because it requires minimal planning and can be reformatted in any plane (2,3). However, such scans lack the temporal information needed to assess cardiac function. A more optimal solution would be a time-resolved (cine) 3D technique. A fundamental problem with this approach is the length of acquisition and the attendant difficulties with respiratory compensation. Parallel imaging techniques (4,5) and other undersampled techniques (6) have been used to acquire 3D cine data sets in a single breath-hold (7-10). However, it is desirable to achieve better combinations of spatiotemporal resolution for an accurate functional and anatomical analysis. Respiratory compensation for static 3D whole-heart imaging can be achieved with the use of navigator beams (11). Unfortunately, navigator techniques interrupt the acquisition and thus are difficult to combine with steady-state free precession (SSFP) cine sequences (12). Furthermore, interleaving of navigators in a cine acquisition can limit the temporal resolution.To address these limitations, it would be useful to develop 3D acquisition techniques that incorporate respiratory self-gating. Such techniques would enable improvements in both spatial and temporal resolution. Respiratory self-navigated techniques have been proposed for 2D radial cine MRI (13), 3D whole-heart coronary MR angiography (MRA) (14) using radial trajectories, and 2D multislice spiral imaging (15). A recent study performed respiratory self-navigated coronary MRA (16) using Cartesian trajectories with a projection calculated from a center k-space profile. In such studies motion compensation is performed retrospectively. The main problem with retrospective correction schemes is that it is difficult to ensure that all necessary data are acquired at the correct respiratory position. It would therefore be preferable to use respiratory self-navigation in a real-time manner such that data corrupted b...

Four-dimensional (4D) flow imaging has been used to study flow patterns and pathophysiology, usually focused on specific thoracic vessels and cardiac chambers. Whole-heart 4D flow at high measurement accuracy covering the entire thoracic cardiovascular system would be desirable to simplify and improve hemodynamic assessment. This has been a challenge because compensation of respiratory motion is difficult to achieve, but it is paramount to limit artifacts and improve accuracy. In this work we propose a self-gating technique for respiratory motion-compensation integrated into a whole-heart 4D flow acquisition that overcomes these challenges. Key words: 4D flow imaging; whole heart; self-gating; Hemodynamic assessment; congenital heart disease Over the past two decades, flow imaging by phase-contrast magnetic resonance imaging (PC-MRI) has become a highly valuable diagnostic method in cardiology, as it allows reliably quantification of blood flow rates and qualitative delineation of flow patterns in the cardiovascular system (1). This is usually performed by the acquisition of a number of two-dimensional (2D) slices with either inplane or through-plane velocity-encoding (VENC). This method is time-consuming and needs careful planning of the various image planes and prior knowledge of both the flow-encoding direction and expected flow velocities (2).Therefore, it usually requires a highly skilled operator, particularly in patients with complicated anatomy such as with congenital heart disease (CHD). It would be much easier to acquire a four-dimensional (4D) MR flow data set covering the entire thoracic cardiovascular system (3). In 4D flow, anatomical and three-directional velocity information are acquired for each pixel within a three dimensional (3D) volume over time (4,5). Flow information can then be analyzed during postprocessing from any reformatted image plane, and prior knowledge of flow directions is not necessary.In the clinical setting, 4D MR flow imaging has mainly been applied to assess flow patterns qualitatively in order to study the pathophysiology of various pathological conditions of the great arteries and cardiac chambers (6 -9), including Fontan-type palliation of complex CHD (10). These studies were usually focused on a particular thoracic vessel or chamber of the heart (11).However, there is much less evidence that 4D flow can reproducibly provide quantitative measures such as stroke volumes (SVs) in various thoracic vessels at a level of accuracy as reached by conventional 2D PC-MR flow. A major concern is potential measurement bias introduced by respiratory motion due to the long acquisition time, constituting the need for effective respiratory motion-compensation strategies. One approach is to obtain multiple signal averages during free-breathing, which has been validated for 2D through-plane flow measurements (2). However, this approach may result in reduced image quality, for which 4D flow is more susceptible than 2D flow as it involves a larger volume and flow directions in three orthogo...