Haemodynamic optimisation of a dialysis graft design using a global optimisation approach

Area of Science:

• Biomedical Engineering
• Fluid Dynamics
• Medical Device Design

Background:

• Arteriovenous graft (AVG) patency is often limited by disturbed blood flow and non-physiological wall shear stress (WSS) at the anastomosis.
• Helical graft modifications are explored to improve hemodynamics and prolong AVG lifespan.
• Traditional hemodynamic optimization relies heavily on computationally intensive CFD simulations.

Purpose of the Study:

• To develop a computationally efficient meta-modeling approach for optimizing AVG geometry.
• To identify an AVG design that minimizes detrimental hemodynamic factors contributing to graft dysfunction.
• To reduce the computational cost associated with hemodynamic optimization of AVGs.

Main Methods:

• A meta-modeling approach was developed using a training dataset of 103 helical graft designs evaluated by CFD.
• Meta-models replaced full CFD simulations during the optimization process.
• The optimized helical graft design was verified using a final CFD simulation.

Main Results:

• The meta-modeling approach reduced the number of CFD evaluations by approximately 2000-fold.
• An optimized helical graft design featuring a helical centerine and helical ridge was identified.
• The optimized design successfully reduced flow disturbances and areas of non-physiological WSS compared to a straight graft.

Conclusions:

• An efficient meta-modeling technique successfully identified an optimized helical AVG design at significantly reduced computational cost.
• The developed helical graft design demonstrates potential for improving AVG patency by mitigating adverse hemodynamic conditions.
• Further in vivo studies are warranted to confirm the clinical benefits of this optimized graft design.