Use of a flexible optical fibre bundle to interrogate a Fabry-Perot sensor for photoacoustic imaging

Abstract

Photoacoustic imaging systems based on a Fabry Perot (FP) ultrasound sensor that is read-out by scanning a free-space laser beam over its surface can provide high resolution photoacoustic images. However, this type of free-space scanning usually requires a bulky 2-axis galvanometer based scanner that is not conducive to the realization of a lightweight compact imaging head. It is also unsuitable for endoscopic applications that may require complex and flexible access. To address these limitations, the use of a flexible, coherent fibre bundle to interrogate the FP sensor has been investigated. A laboratory set-up comprising an x-y scanner, a commercially available, 1.35 mm diameter, 18,000 core flexible fibre bundle with a custom-designed telecentric optical relay at its distal end was used. Measurements of the optical and acoustic performance of the FP sensor were made and compared to that obtained using a conventional free-space FP based scanner. Spatial variations in acoustic sensitivity were greater and the SNR lower with the fibre bundle implementation but high quality photoacoustic images could still be obtained. 3D images of phantoms and ex vivo tissues with a spatial resolution and fidelity consistent with a free-space scanner were acquired. By demonstrating the feasibility of interrogating the FP sensor with a flexible fibre bundle, this study advances the realization of compact hand-held clinical scanners and flexible endoscopic devices based on the FP sensing concept.

Fig. 1.

Experimental set-up illustrating the interrogation of the FP sensor using a flexible fibre bundle and telecentric lens relay system comprising ${\mathrm{L}}_1$ and ${\mathrm{L}}_2$.

Fig. 2.

Relative round-trip coupling efficiency. (a) Scanned image of the fibre bundle using the set-up shown in Fig. 1. The greyscale intensity represents the measured photodiode voltage, ${\mathrm{V}}_{\mathrm{d}\mathrm{c}}$, and provides a measure of the relative round-trip coupling efficiency of the system. (b) Histogram of data in (a) (${\mathrm{V}}_{\mathrm{d}\mathrm{c}}$ corresponds to the core centres) and that acquired from an identical scan of the same FP sensor over the same area and number of spatial points (18,000) using the free space FP scanner. The vertical axis in (b) represents the number of spatial points of value ${\mathrm{V}}_{\mathrm{d}\mathrm{c}}$ expressed as a percentage of the total number of points scanned (18,000).

Fig. 3.

Interferometer transfer function (ITF) of the FP sensor interrogated by a single core of the fibre bundle (orange) and its Lorentzian fit (gray). The ITF of the same sensor interrogated by a free-space beam is also shown (blue) for comparison.

Fig. 4.

Histograms of (a) signal, (b) noise and (c) signal-to-noise ratio (SNR) for fibre bundle and free-space interrogated FP sensor configurations. In both cases the FP sensor was interrogated at 18,000 different points over a circular area of 10 mm diameter. The vertical axes represents the number of spatial points expressed as a percentage of the total number of points scanned (18,000).

Fig. 5.

Spatial resolution of the fibre bundle FP sensor imaging system. (a) Reconstructed PA image showing ribbon cross-sections at different depths. (b) Lateral and (c) axial profiles through the ribbon feature identified by the dotted red rectangle in (a), respectively. (d) A contour plot showing the lateral spatial resolution in the x-z plane.

Fig. 6.

PA images (6 mm aperture) of arbitrary shaped phantoms. Top panel: widefield microscope images of a synthetic hair knot and leaf skeleton phantom coated in India ink. Middle and lower panels: reconstructed PA images of the phantoms shown as maximum intensity projected along the x-y and x-z planes.

Fig. 7.

PA images (10 mm aperture) of an ex vivo duck chorioallantoic membrane (CAM) where microvasculature is clearly visualized. The images are-coded according to the depth and maximum intensity projected along the x-y and x-z planes. Laser excitation wavelength: 590 nm, fluence: 18 mJ cm⁻².

Fig. 8.

PA images (10 mm aperture) acquired at three locations on an ex vivo term normal human placenta. Top panel: widefield microscope images from the fetal side of the placenta where chorionic (fetal) vessels are visualised. Middle and lower panels: 3D PA images of the same area as top panel,-coded according to the depth and maximum intensity projected along the x-y and x-z planes. Locations marked by letter v and c indicate areas where sub-surface chorionic vessels and calcium deposits are visualized, respectively. Laser excitation wavelength: 590 nm, fluence : 18 mJ cm⁻².

Fig. 9.

(left) Ray diagram illustrating the relationship between angle of the wedge and angles at which the fibre bundle endface is polished and tilted to suppress the detection of Fresnel reflections. (right) Histogram plot shows RMS noise floor of the FP sensing elements before (blue) and after (orange) angle polishing and wedging of the bundle endfaces. The noise was measured by scanning 3,000 cores in the centre of the bundle with 20 MHz detection bandwidth.