- Conference Article
- 10.1109/ius62464.2025.11201693
DOPCON: A DOuble-Pass CONvolution beamformer for High-Resolution Imaging in Handheld Ultrasound Systems
- Sep 15, 2025
- Barnali Kundu + 2 more +2
High-resolution ultrasound imaging with lower computational complexity remains a key research objective in medical diagnostics. Linear beamforming techniques such as DAS are attractive for their computational efficiency, but they are fundamentally limited by poor resolution and low contrast Nonlinear beamformers, like DMAS and COBA, have shown to outperform DAS by providing improved resolution and contrast, thereby enabling enhanced visualization of deeper tissue structures. More recently, the double-stage DMAS (DS-DMAS) algorithm has demonstrated further improvements over DMAS, albeit with a computational complexity of $\mathcal{O}\left( {{M^2}} \right)$. In this work, we introduce DOPCON (DOuble-Pass CONvolution), a novel nonlinear beamformer that achieves superior image quality compared to DAS, DMAS, COBA, and DS-DMAS, while reducing the computational complexity to $\mathcal{O}(M\log M)$. The proposed formulation also incorporates apodization, further enhancing image quality metrics. Field-II simulation studies demonstrate that DOPCON offers improved sidelobe suppression, contrast, and resolution. Furthermore, we present an optimized variant of DOPCON that reduces one of the key computational block from $\mathcal{O}(M\log M)$ to $\mathcal{O}(M)$, making the approach particularly well suited for resource-constrained handheld ultrasound systems with high-resolution imaging.
Read more