A General Approach for Optimization of Ultrasonic Array Parameter and Setup Using Principal Component Regression.

Area of Science:

• Engineering
• Materials Science
• Medical Imaging

Background:

• Ultrasonic array design requires understanding the interplay between acoustic beam characteristics (ABCs) and imaging performance.
• Existing methods may not efficiently link these critical design parameters for optimization.

Purpose of the Study:

• To propose a general optimization approach for ultrasonic array design.
• To establish a quantitative relationship between ABCs and imaging performance indicators (IPIs).
• To enable simultaneous optimization of beam and imaging performance.

Main Methods:

• Utilized principal component regression (PCR) to model the relationship between ABCs and IPIs.
• Developed an optimization model using the PCR-derived mapping function as the objective function.
• Employed a global optimization algorithm to find optimal array parameters and configurations.

Main Results:

• Demonstrated the effectiveness of the PCR-based approach in linear and circular ultrasonic array designs.
• Quantitatively established the association between acoustic beam and imaging performance through PCR mapping.
• Showcased the ability to optimize beam and imaging performance concurrently, skipping acoustic response modeling during iterations.

Conclusions:

• The proposed PCR-based optimization approach provides an interpretable and explicit link between ultrasonic array design parameters and performance.
• This method facilitates simultaneous optimization of acoustic beam characteristics and imaging performance.
• The approach accelerates the design process by potentially bypassing computationally intensive acoustic response modeling.