Virtual Cardiac Surgery Planning to Improve Blood Flow Dynamics using 3D Image Processing
Chin Siang Ong1, Yue-Hin Loke2, Dominik Siallagan3, Laura Olivieri2, Diane de Zélicourt4, William Moore3, Marianne Schmid Daners5, Mirko Meboldt5, Luca Vricella1, Axel Krieger3, Narutoshi Hibino1
1Division of Cardiac Surgery, Johns Hopkins Hospital, Baltimore, MD;2Division of Cardiology, Children's National Health System, Washington DC, DC;3Sheikh Zayed Institute for Pediatric Surgical Innovation, Children's National Health System, Washington DC, DC;4Institute of Physiology, University of Zürich, Zürich, Switzerland5Institute of Design, Materials and Fabrication, Eidgenössische Technische Hochschule Zürich, Zürich, Switzerland
Objective: Despite advances in the surgical management of congenital heart disease, significant morbidity and mortality arise due to the complexity of surgery, secondary to diverse patient anatomies. Ideal patient-specific conduit design is important for the success of surgical reconstructive repair for hypoplastic vessels or stenosis, and to avoid hemodynamically significant residual disease. However, current standard preoperative preparation does not include patient-specific conduit design. Preoperative patient-specific conduit design for the most ideal reconstructive route with the most optimal flow dynamics may yield long-term benefits for patients' health and quality of life. The objective of this study is to determine if unique, patient-specific three-dimensional (3D) virtual cardiac surgery models will allow congenital heart surgeons to preoperatively visualize the Fontan circuit and perform patient-specific conduit design, in order to ensure the most optimal flow dynamics and anatomical arrangement, as an in silico rehearsal for the actual surgery.
Methods: Cross-sectional images of patients who underwent Fontan surgery (n=2) were used to create 3D models of Fontan circuits using standard segmentation software. For each patient, three 3D models were created, "native" (the patient's actual post-operative Fontan circuit), a virtually designed "flared end" conduit and a virtually designed "bifurcated Y" conduit (Figure 1), with the goal of improving hepatic flow distribution (HFD) and reducing power loss. Steady-state computational fluid dynamic (CFD) simulations were performed on all six models, using MR-derived flow splits for boundary conditions to assess HFD and power loss.
Results: Two types of virtually designed conduits ("flared end" and "bifurcated Y") were successfully created for each Fontan patient, accommodating the conduit insertion sites, the size, length, direction and angle of their respective Fontan routes. CFD simulations showed that the "bifurcated Y" grafts improved HFD between left and right pulmonary artery from 77/23 to 63/37 and from 35/65 to 47/53, respectively (Figure 1). Both "flared end" and "bifurcated Y" graft reduced power loss from 8.67 mW (native) to 6.43 mW and 6.93 mW (Patient 1), and 11.28 mW (native) to 8.09 mW and 4.66 mW (Patient 2) respectively.
Conclusions: Our virtual cardiac surgery approach has the potential to improve the quality of surgery by designing the most optimal conduit design before surgery, that has the most optimal HFD and least power loss. Future work will include validation of CFD calculations and further optimization of the conduit design model, using MRI data from more patients.
Back to 2017 Program