Related Experiment Video
Updated: Jan 6, 2026
04:55
Author Spotlight: Simulating Pediatric Cardiac Surgery Using a Neonatal Piglet Model
Published on: May 26, 2023
1.3K
A Patient-Specific Computational Model for Neonates and Infants with Borderline Left Ventricles
Yurui Chen1, Isao A Anzai2, David M Kalfa3,4
1Department of Mechanical Engineering, Columbia University, 500 W 120th Street, MC 4703, New York, NY, 10027, USA.
Annals of Biomedical Engineering
|November 4, 2025
Summary
Computational modeling predicts surgical outcomes for borderline left ventricle (BLV) patients. This tool helps surgeons choose between biventricular repair (BiVR) and Stage 1 palliation (S1P), improving decision-making for BLV infants.
Area of Science:
- Computational modeling in pediatric cardiology
- Hemodynamic analysis of congenital heart disease
Background:
- Borderline left ventricle (BLV) poses a surgical dilemma between biventricular repair (BiVR) and Stage 1 palliation (S1P).
- Choosing BiVR for BLV can increase mortality risk due to potential hemodynamic instability.
- Accurate prediction of surgical outcomes is crucial for optimizing treatment strategies in BLV patients.
Purpose of the Study:
- To develop and validate a personalized computational model for predicting surgical hemodynamics in BLV patients.
- To assist surgical decision-making by simulating virtual surgeries (BiVR and S1P) and quantifying hemodynamic outcomes.
- To provide a predictive tool for determining the hemodynamically optimal surgical approach for BLV infants.
Main Methods:
- Developed a novel multi-block lumped parameter network (LPN) model of the BLV circulatory system.
- Estimated patient-specific model parameters using a semi-automatic tuning framework with clinical data from ten BLV patients.
- Performed virtual BiVR and S1P surgeries on each patient model to assess post-operative hemodynamics.
Main Results:
- In patients who underwent S1P, virtual BiVR predicted significantly elevated mean pulmonary artery pressure (PAPmean), mean left atrial pressure (LAPmean), and single-ventricle end-diastolic pressure (SVEDP).
- Virtual BiVR in patients who underwent BiVR did not predict adverse hemodynamic outcomes.
- Model predictions aligned with clinically adopted procedures, highlighting potential risks of BiVR in certain BLV cases.
Conclusions:
- A novel subject-specific computational modeling framework can predict hemodynamics following virtual surgery in BLV patients.
- The model accurately predicted adverse hemodynamic outcomes for virtual BiVR in patients who clinically received S1P.
- This predictive tool shows promise for guiding surgical decisions in BLV infants, requiring prospective validation.
