Enhanced resolution optoacoustic microscopy using a picosecond high repetition rate Q-switched microchip laser

Abstract

Conventional optoacoustic microscopy (OAM) instruments have at their core a nanosecond pulse duration laser. If lasers with a shorter pulse duration are used, broader, higher frequency ultrasound waves are expected to be generated and as a result, the axial resolution of the instrument is improved. Here, we exploit the advantage offered by a picosecond duration pulse laser to enhance the axial resolution of an OAM instrument. In comparison to an instrument equipped with a 2-ns pulse duration laser, an improvement in the axial resolution of 50% is experimentally demonstrated by using excitation pulses of only 85 ps. To illustrate the capability of the instrument to generate high-quality optoacoustic images, en-face, in-vivo images of the brain of Xenopus laevis tadpole are presented with a lateral resolution of $3.8\mu m$ throughout the entire axial imaging range.

Keywords: Q-switched lasers; axial resolution enhancement; optoacoustic.

Fig. 1

(a) Schematic diagram. OS1: picosecond laser; OS2: supercontinuum optical source; C1: reflective collimator; SH: sample holder; FM: flipping mirror; PD: photodetector; GS: orthogonal galvo-scanners; DAQ1,2: data acquisition cards; LNA: low noise amplifiers; UT: ultrasound transducer; OL: objective lens. TS: translation stage; TTL1,2: TTL signals synchronized with the emission of the pulses. (b) Picture showing the UT and SH.

Fig. 2

(a) Experimentally measured edge (magenta) and line (green) spread functions. (b) Detected acoustic signal versus axial position (data provided by the manufacturer of the transducer). (c) Lateral FOV, measured by imaging a carbon fiber tape.

Fig. 3

Typical OA signals generated by exciting a carbon fiber with a 2-ns pulse duration [green curve in (a)] and 85-ps pulse duration laser [red curve in (b)]. The envelopes of the two signals are presented in pink and blue, respectively. From the signals presented in (a) and (b), the acoustic spectra generated by using OS1 and OS2 were calculated in (c). By measuring the FWHM of the two spectra, we could infer axial resolutions of 25 and $51\mu \mathrm{m}$, respectively. The fact that the two spectra are not identical in terms of central frequency and bandwidth shows that, the axial resolution is not determined by the bandwidth of the transducer alone.

Fig. 4

Part A: flowchart describing the imaging protocol and the post-processing steps. Part B: (a) microscope image of the tadpole’s head over which the OAM image showed in (b) is overlapped. (b) Composite en-face image obtained by merging images collected at 24 axial positions separated by $50\mu \mathrm{m}$. (c)–(e) Single-plane images showing significant examples of defined brain structures that appear as the focal plane is shifted deeper into the tadpole. The axial separation between (c) and (d), and (d) and (e) is $200\mu \mathrm{m}$. It is noteworthy that (a) has the same lateral size as (b) and (c) has the same lateral size as (d) and (e). Part C: (f) and (g) B-scan images of the carbon tape produced using the ps and the ns lasers, respectively. Axial size (along the horizontal direction): 1.6 mm. Lateral size (vertical direction): $50\mu \mathrm{m}$. (h) and (i), typical en-face images of the tadpole’s eye produced using the ps and the ns lasers, respectively. In both cases, the light is focused inside the eye. The artifact in (i) is due to a structural defect of the optical window.