Uniform-Angle Sampling and Reconstruction Enhancement in Compact PLD-MEMS Photoacoustic Microscopy

Abstract

Compact photoacoustic microscopy (PAM) systems based on pulsed laser diodes (PLDs) and micro-electro-mechanical system (MEMS) scanners offer reduced system size and cost, but typically operate under limited excitation energy and nonuniform spatial sampling. The resonant motion of MEMS scanners leads to spatially varying sampling density, while the low pulse energy of PLDs constraints the achievable signal-to-noise ratio (SNR) and acquisition throughput. In this work, PAM is implemented using PLD excitation and a two-axis MEMS scanning mirror. To improve spatial sampling consistency under resonant MEMS motion, a uniform-angle multicycle sampling strategy is introduced. In addition, a learning-based signal restoration method is employed to enhance reconstruction robustness under low-SNR acquisition conditions. Experimental validation using vascular phantoms and in vivo mouse ear imaging demonstrates that fine absorptive structures can be reliably resolved with as few as 20 excitations. The combined use of sampling optimization and signal restoration reduces sampling-induced distortion and the dependence on extensive signal averaging, enabling efficient PAM under joint low-SNR and nonuniform sampling constraints.