Max-Divergence Tri-Angle GAN: Optimizing ultrasound plane wave image quality with strategic angle selection and attention mechanism

Plane-wave imaging (PWI) offers high frame rates but often suffers from reduced image quality. To improve this, multiple angles are typically compounded, but this comes at the cost of temporal resolution. To address this trade-off, we introduce the Max-Divergence Tri-Angle Generative Adversarial Network (Tri-AngleGAN), which strategically selects and exploits three plane wave angles with the greatest divergence—leftmost, rightmost, and center. An attention mechanism is integrated to effectively harness and correlate the information from these angles, maximizing their content. This approach enables efficient reconstruction of high-resolution images and enhances overall imaging quality with fewer inputs, providing an advanced solution for improving image quality in PWI.