Acoustic coherence-aware oriented 3D photoacoustic microscopy super-resolution with noise-suppression

Photoacoustic microscopy (PAM) balances optical contrast with ultrasound resolution but faces trade-offs among laser safety, imaging speed, and signal quality. Conventional 2D-based approaches discard axial acoustic coherence, causing irreversible spatial information loss. To break through this bottleneck, we propose an acoustic coherence-aware super-resolution and noise-suppression network (ACORNet) that directly reconstructs raw 3D photoacoustic data. Under severe conditions combining spatial down-sampling (up to ×8) with heavy noise, ACORNet robustly preserves 3D microvascular continuity and high-frequency boundaries. Furthermore, it enables reliable dual-wavelength functional oxygen saturation (sO₂) imaging, providing an effective solution for safe, high-speed in vivo computational microscopy.