Young Investigator Award session – Basic Science3433D printed models for surgical planning in complex congenital heart disease344Ultrafast doppler imaging of intramyocardial coronary arteries345Quantification of mitral regurgitation with multiple jets: in vitro comparison of two-dimensional PISA techniques346Non-invasive ultrasonic chordal cutting

343 3D printed models for surgical planning in complex congenital heart disease {#article-title-2} Purpose: To explore the use of 3D patient-specific cardiovascular models using rapid prototyping techniques to avoid unexpected findings and improve surgical planning in patients with complex congenital heart disease Methods: This European prospective multicenter (7 institutions) study included 21 patients with complex congenital heart diseases. Magnetic resonance imaging (MRI) and computed tomography (CT) were used to acquire 3D cardiovascular anatomy. Fused deposition technique using polylactic acid was used.A Bland-Altman analysis was used to evaluate the diameters measurement agreement between the 3D printed model and the patient's MRI and CT. 3D-models were used to plan the surgery.After the procedure, surgeons involved filled a questionnaire form to evaluate the usefulness of the 3D printed model. Results: The Bland-Altman analysis showed accurate agreement in the diameter between medical images and 3D-models (-0.20 ± 1.70 mm). 3D-models showed the spatial relationships between the VSD and great vessels in 10 patients with DORV and 3 patients with TGA and pulmonary stenosis, re-appraisal for biventricular repair in 2 cases, planning of lateral tunnel Fontan completion in 2 cases, re-opening of a restrictive VSD in 1 case, evaluation of MAPCA's in 1 case and evaluation of criss-cross heart anatomy in 2 cases. Surgeons found the 3D models to be very useful for surgical planning and avoiding complications with an overall level of satisfaction of 8.5 out of 10. Conclusions: 3D-printed cardiovascular models accurately replicate the patient's anatomy and are extremely helpful for planning surgery in complex congenital heart disease. They may potentially reduce operative time and morbi-mortality. ![Figure][1] # 344 Ultrafast doppler imaging of intramyocardial coronary arteries {#article-title-3} Purpose: Transthoracic echocardiography is recommended by international guidelines to assess wall motion abnormalities suggestive of coronary artery disease (CAD). As conventional Doppler imaging is not sensitive enough to assess myocardial perfusion, the risk stratification of CAD patients can be enhanced by performing myocardial contrast echocardiography (MCE). Unfortunately, MCE is scarcely used in clinical practice. Here, we implemented ultrafast Doppler imaging combined with cardiac motion correction on a conventional 3 MHz cardiac phased array probe and successfully imaged intramyocardial coronary arteries without any contrast agent injection. In addition, the sensitivity of ultrafast Doppler imaging to changes in myocardial perfusion was positively assessed by measuring Doppler intensity in the myocardium during occlusion and reperfusion of the left circumflex (LCx) and left anterior descending (LAD) coronary territories. Methods: We used a 3 MHz phased array probe connected to a programmable ultrafast ultrasound scanner (Aixplorer, Supersonic Imagine, France). We insonified open chest pig hearts (N=3) in diastole with a dedicated ultrafast Doppler imaging sequence consisting of 8 diverging wave transmissions at a pulse repetition frequency of 1 kHz. Electrocardiograms and coronary volumetric flow rates measured invasively (ultrasonic Doppler flow meter) were registered during ultrasound acquisitions. Ultrafast datasets were beamformed and tissue clutter was separated from coronary signals with a sliding multidimensional filter based on spatiotemporal coherence analysis. Ultrafast Power Doppler images of intramyocardial coronary arteries were computed and overlaid on ultrafast Bmode images to cross anatomical and functional information. Results: We successfully imaged intramyocardial coronary arteries in early and late diastole. Time windows (< 100 ms) for coronary imaging corresponded to phases of the heart cycle where wall motion does not exceed 2 cm/s as assessed by Tissue Doppler. Ultrafast Doppler intensity changes in the myocardium during LCx and LAD occlusion and subsequent reperfusion were of the order of 70 % which correlated positively with coronary flow rates measured invasively, whereas no significant drop was observed in Bmode intensities. Intramyocardial artery cross sections measured with ultrafast Doppler imaging went down to 150 μm. Conclusion: Thanks to its very high sensitivity, ultrafast Doppler imaging of intramyocardial coronary arteries demonstrates strong clinical potential as an ultrasound-based coronarography technique free of any contrast agent injection. # 345 Quantification of mitral regurgitation with multiple jets: in vitro comparison of two-dimensional PISA techniques {#article-title-4} Purpose: Traditionally, the proximal isovelocity surface area (PISA) is based on the assumption of a single hemisphere (hemispheric PISA), but this technique has not been validated for the quantification of mitral regurgitation (MR) with multiple jets. Methods: The left heart simulator was actuated by a pulsatile pump at various stroke amplitudes. The regurgitant volume (Rvol) passing through the mitral valve phantoms with single and double regurgitant orifices of varying size and interspace was quantified by a flowmeter as reference technique. Color Doppler 3-D full-volumes were obtained, and Rvol were derived from 2-D PISA surfaces on the basis of hemispheric and hemicylindric assumption with one base (partial hemicylindric PISA) or 2 bases (total hemicylindric PISA). Results: 72 regurgitant volumes (Rvol range: 8 to 76 ml/beat) were obtained. Hemispheric PISA Rvol correlated well with reference Rvol by one orifice (R²=0.97; bias -2.7 ± 3.2ml), but less by ≥ one orifice (R²=0.89). When a fusion of two PISAs occured, addition of two hemispheric PISA overestimated Rvol (bias 9.1 ± 12.2ml, fig.1), and single hemispheric PISA underestimated Rvol (bias -12.4 ± 4.9ml). If an integrated approach was used (hemispheric in single orifice, total hemicylindric in two non-fused PISAs and partial hemicylindric in two fused PISAs), the correlation was R²=0.95, bias -1.6 ± 5.6ml (fig.2). In the ROC analysis, the cutoff to detect ≥ moderate-to-severe Rvol (≥45ml) was 42ml (AUC 0.99, sens. 100%, spec. 93%). Conclusions: In MR with two regurgitant jets, the 2-D hemicylindric assumption of the PISA offers a better quantification of Rvol than the hemispheric assumption. Quantification of MR using 2-D PISA requires an integrated approach that considers number of regurgitant orifices and fusion of the PISAs. ![Figure][1] Abstract 90816 Figures 1 and 2. # 346 Non-invasive ultrasonic chordal cutting {#article-title-5} Objective: Basal chordae surgical section has been shown to be effective to reduce mitral regurgitation (MR). We investigated in vitro and in vivo the feasibility of pulsed cavitational focused ultrasound (histotripsy) for non-invasive chordal cutting to avoid cardiopulmonary bypass and invasive surgery in infarcted heart. Methods: Experiments were performed in vitro in explanted sheep hearts (N=10) and in vivo in beating sheep hearts (N=7). In vitro, the mitral valve (MV) apparatus including basal chordae was removed, fixed on a holder in a water tank. High intensity ultrasound pulses were emitted from the therapeutic device (1-MHz focused transducer) placed at a distance of 64 mm under echocardiography guidance. In vivo, after sternotomy, the same therapeutic device was applied on the beating heart. We analyzed mitral valve coaptation and chordae by real time 3D echocardiography before and after basal chordal cutting. After sacrifice, anatomical and histological postmortem explorations were realized. Results: In vitro, all basal chordae were completely cut after mean procedure duration of 5.5 ± 2.5 minutes. The duration of the procedure was found to increase linearly with the chordae diameter. In vivo, the central basal chorda of the anterior leaflet were completely cut. The mean procedure duration was 20 ± 9 minutes (min=14; max=26). The sectioned chorda was visible on echocardiography and MV coaptation remained normal with no significant MR. Anatomical and histological postmortem explorations of hearts confirmed the section of the chordae. Conclusions: Histotripsy achieved successfully to cut mitral valve chordae in vitro and in vivo in beating heart. This technique can open the door to the non-invasive treatment of functional mitral regurgitation. ![Figure][1] Chordal cutting by histotripsy [1]: pending:yes