Multi-modality photoacoustic microscopy (mPAM) is a powerful technique for revealing disease pathology, yet its optical-resolution mode is fundamentally limited by the shallow depth-of-field (DoF) of Gaussian beams. Here, we proposed a decision-level constrained end-to-end multi-focus image fusion (Dc-EEMF) powered mPAM for extension of DoF. By integrating a lightweight siamese network with an artifact-resistant channel-wise spatial frequency fusion rule and a U-Net-based decision level focus perceptual loss, Dc-EEMF effectively combines the advantages of spatial and transform domain fusion strategies. It not only achieves superior fusion performance within a single modality, outperforming existing fusion techniques, but also demonstrates strong generalizability across diverse microscopy platforms.
Zhou et al. (Thu,) studied this question.