Fast laser diode scanning photoacoustic microscopy using direct model inversion
Abstract
In optical-resolution photoacoustic microscopy (OR-PAM), laser scanning allows for fast imaging compared to mechanical confocal raster scanning of the focused ultrasound sensor. However, the image field-of-view (FOV) is restricted to the high-sensitivity acoustic focus region. To overcome this, we propose a calibrated spatial impulse response (SIR)-based algorithm using a fast, direct model inversion (DiMI) to enlarge the FOV beyond the acoustic focus. DiMI has been implemented in a cost-effective reflection mode OR-PAM system based on a laser diode (LD). Despite their low-cost and compactness, LDs emit low-energy laser beams, necessitating tightly focused optics and transducers with short working distances, which restricts the FOV to an even narrower acoustic focus. Enabled by DiMI, we present the first reflection mode laser scanning LD-based PAM microscope. This method significantly expands the effective FOV by two orders of magnitude beyond the acoustic focus of 150 μm with homogeneous SNR (> 20 dB), acquiring 1x1 mm2 in less than 5 seconds compared to more than 10 minutes using raster-scanning. Additionally, to obtain volumetric images of up to 1 mm, we perform vertical depth scans of the microscope head. DiMI enhances the limited axial resolution of the acquired 3D images by 43% compared to confocal raster-scanning. This novel approach enabled accurate volumetric imaging of phantoms and melanocytic lesions in murine models in vivo, with strong histological agreement of the tumor aspect ratio. These results demonstrate a new pathway for implementing fast-scanning mirrors in clinical-oriented OR-PAM systems, enhancing the translation of practical, cost-effective photoacoustic imaging.
Fields of Study & Subjects
Photoacoustic and Ultrasonic Imaging
Thermography and Photoacoustic Techniques
Spectroscopy Techniques in Biomedical and Chemical Research
Publication details
- Journal / source
- Photons Plus Ultrasound: Imaging and Sensing
- Pages
- 10–10
- Publication date
- 4 March 2026
- Language
- EN