An Efficient Way to Find Optimal Crossover Designs Using CVX for Precision Medicine.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Precision Medicine

Background:

• Crossover designs are increasingly vital in precision medicine.
• Optimal design is essential for efficient and reliable clinical trials.

Purpose of the Study:

• To formulate the optimal crossover design problem as a convex optimization problem.
• To demonstrate the effectiveness of the CVX package for finding optimal crossover designs.
• To explore N-of-1 and dual-objective optimal crossover designs for precision medicine applications.

Main Methods:

• Formulation of the optimal crossover design problem as a convex optimization problem.
• Utilizing the CVX optimization package to search for optimal designs.
• Application of CVX for N-of-1 and dual-objective crossover designs.
• Illustrating the method with A-optimality criterion.

Main Results:

• The CVX package is effective for finding optimal crossover designs, even when analytical solutions are unavailable.
• CVX can identify optimal designs for N-of-1 trials and dual-objective crossover designs.
• The study provides CVX codes for practical application.

Conclusions:

• Convex optimization using CVX is a powerful tool for discovering optimal crossover designs in precision medicine.
• This approach facilitates the search for complex designs, improving treatment and carryover effect estimation.
• The provided codes enable broader application of these advanced design strategies.