The multi-response linear polynomial model is transformed into a single response polynomial model. In symmetric test area, for the polynomial models the A-D- and E-optimal experimental designs are obtained.

Transcatheter pulmonary valve replacement (TPVR) is a technique for treating valvular heart diseases or dysfunctions without the need for open-heart surgery. Commercial valve stents, such as Epic TM valved stents or mechanical heart valves, can be used to solve the problems of narrowed or leaky pulmonary valves and also to improve regurgitation flow and heart pump efficiency (HPE). However, these prosthetic valves have limitations in terms of availability, durability, and stent size for children and for meeting the needs of special subjects. To be deemed fit for new clinical cases, handmade trileaflet-valved conduits (HTVCs) provide a promising strategy for designing stents using optimal design methods for customized specifications for child and adult patients. Before clinical applications, the functions of HTVCs need to verify hemodynamic statuses, such as the regurgitation fraction (RF), HPE, and pressure drop, under different heart rates (HR), blood flow volumes, and hypertension conditions. To reduce the number of experimental tests, the Taguchi method with signal-tonoise (S/N) ratios is used to validate the assigned optimal parameter designs, which can obtain good hemodynamic performance with the desired target goals of (1) RF ≤ 20% (minimizing the objective function) and (2) HPE ≥ 80% (maximizing the objective function). In contrast to commercial valve stents, such as Epic TM valved stent and mechanical heart valve, likelihood degree estimation is also employed to quantify the degree of valve performance and verify the quality of HTVCs.