🤖 AI Summary
To address the inherent trade-off between high scanning speed and degraded spatial resolution in rotational-scanning photoacoustic microscopy (PAM), this paper proposes the first super-resolution reconstruction framework specifically designed for this modality. Methodologically, we: (i) formulate a rotation-scanning-specific motion degradation model; (ii) design an odd-even line registration module to correct non-uniform mechanical deformations induced by rotation; (iii) introduce a gradient-driven, plaque-adaptive sampling strategy that prioritizes critical vascular structures; and (iv) incorporate a Transformer architecture to model long-range dependencies and fuse global contextual information. Extensive experiments on both synthetic and real PAM datasets demonstrate significant improvements in PSNR and SSIM, with markedly enhanced recovery of fine, continuous vascular details. The source code is publicly available.
📝 Abstract
Photoacoustic Microscopy (PAM) images integrating the advantages of optical contrast and acoustic resolution have been widely used in brain studies. However, there exists a trade-off between scanning speed and image resolution. Compared with traditional raster scanning, rotational scanning provides good opportunities for fast PAM imaging by optimizing the scanning mechanism. Recently, there is a trend to incorporate deep learning into the scanning process to further increase the scanning speed.Yet, most such attempts are performed for raster scanning while those for rotational scanning are relatively rare. In this study, we propose a novel and well-performing super-resolution framework for rotational scanning-based PAM imaging. To eliminate adjacent rows' displacements due to subject motion or high-frequency scanning distortion,we introduce a registration module across odd and even rows in the preprocessing and incorporate displacement degradation in the training. Besides, gradient-based patch selection is proposed to increase the probability of blood vessel patches being selected for training. A Transformer-based network with a global receptive field is applied for better performance. Experimental results on both synthetic and real datasets demonstrate the effectiveness and generalizability of our proposed framework for rotationally scanned PAM images'super-resolution, both quantitatively and qualitatively. Code is available at https://github.com/11710615/PAMSR.git.
