Heavy water coupling gel for short-wave infrared photoacoustic imaging

Abstract

Significance: Changes in lipid, water, and collagen (LWC) content in tissue are associated with numerous medical abnormalities (cancer, atherosclerosis, and Alzheimer's disease). Standard imaging modalities are limited in resolution, specificity, and/or penetration for quantifying these changes. Short-wave infrared (SWIR) photoacoustic imaging (PAI) has the potential to overcome these challenges by exploiting the unique optical absorption properties of $\mathrm{LWC}>1000\mathrm{nm}$.

Aim: This study's aim is to harness SWIR PAI for mapping LWC changes in tissue. The focus lies in devising a reflection-mode PAI technique that surmounts current limitations related to SWIR light delivery.

Approach: To enhance light delivery for reflection-mode SWIR PAI, we designed a deuterium oxide ($D_{2}O$, "heavy water") gelatin (HWG) interface for opto-acoustic coupling, intended to significantly improve light transmission above 1200 nm.

Results: HWG permits light delivery $>1\mathrm{mJ}$ up to 1850 nm, which was not possible with water-based coupling ($>1\mathrm{mJ}$ light delivery up to 1350 nm). PAI using the HWG interface and the Visualsonics Vevo LAZR-X reveals a signal increase up to 24 dB at 1720 nm in lipid-rich regions.

Conclusions: By overcoming barriers related to light penetration, the HWG coupling interface enables accurate quantification/monitoring of biomarkers like LWC using reflection-mode PAI. This technological stride offers potential for tracking changes in chronic diseases (in vivo) and evaluating their responses to therapeutic interventions.

Keywords: cancer; collagen; heavy water; high resolution ultrasound; lipids; optoacoustic imaging; photoacoustic imaging and spectroscopy; short-wave infrared.

Fig. 1

Optical absorption coefficients for lipid, water, and collagen between 700 and 1800 nm.^–

Fig. 2

SWIR absorption coefficients for water (${\mathrm{H}}_2\mathrm{O}$) and heavy water (${\mathrm{D}}_2\mathrm{O}$) between 1200 and 2000 nm. Half-value propagation distance dramatically decreases above 1200 nm for ${\mathrm{H}}_2\mathrm{O}$ compared to ${\mathrm{D}}_2\mathrm{O}$.

Fig. 3

(a) Transmission-mode set up of the Vevo LAZR-X using the MX250 probe with LFB = laser fiber bundles, MC = 3D printed molding container, CM = coupling medium, BA = broadband absorber/electrical tape, WR = water reservoir, and US= linear array US probe. 2% and 3% w/w HWG samples are illuminated from above, and the resulting PA spectrum of the broadband absorber is used to quantify optical loss through the optical path. Acoustic coupling is achieved via the water reservoir in contact with the US array. (b) Standard reflection-mode setup with the coupling medium inserted between the probe and sample with LI = laser illumination pattern and S = sample.

Fig. 4

(a) Coupling media PA transmission results. Transmission of HWG and WG at 2% and 3% w/w gellan concentrations are displayed, along with ${\mathrm{D}}_2\mathrm{O}$ (100% concentration, pure), ${\mathrm{D}}_2\mathrm{O}$ (99% concentration, 1% ${\mathrm{H}}_2\mathrm{O}$) and ${\mathrm{H}}_2\mathrm{O}$ (100% concentration, pure) calculated transmission via absorption coefficients through an identical pathlength $\mathrm{\Delta }z=10\mathrm{mm}$. (b) Laser energy measurements of the reflection-mode setup through samples of HWG and WG (thickness $\sim 5.0\mathrm{mm}$) with the horizontal line denoting the cut-off energy threshold of 1 mJ.

Fig. 5

Cross sectional PE images at 25 MHz through graphite phantom in different coupling media. The embedded fine graphite particles served as point scatterers for evaluating the acoustic propagation though the different coupling media: humimic medical rubber (black), HWG (blue), and WG (red). The curves represent the axial PSFs with the FWHMs representing the axial resolutions. Green scale bar denotes 1 mm. Starting depth for each B-mode image is 3 mm from the transducer head.

Fig. 6

Reflection-mode PA images of lipid/water phantom (S) at peak absorption wavelengths of 1220 nm (lipid), 1450 nm, (water), and 1720 nm (lipid) through 4.5 mm of WG and HWG coupling. Yellow scale bar is 1 mm in both depth and lateral directions. Dashed green line indicates the coupling/sample boundary. No useful PA images were obtained through WG above $\sim 1350\mathrm{nm}$ due to the strong absorption of water coupling.

Fig. 7

(a) Amplitude of PA signals from the surface of the lipid/water phantom through 5 mm thick HWG and WG coupling; the dark blue plot corrects for wavelength-dependent absorption through HWG (divides raw spectrum by % transmission of HWG at 2% w/w as previously recorded). (b) Published SWIR absorption spectra of lipid and water.^, The cut-off energy for detecting PA signals from the sample was defined as 1 mJ for this study (denoted by the red dotted vertical line). Acceptable wavelengths through WG were 1200 to 1350 nm, and 1200 to 1850 nm through HWG.

Fig. 8

(a) Photograph of bovine muscle sample cross section displaying the approximate slice presented in the PA images. (b) Reflection-mode PA images at absorption peaks of lipid (1220 and 1720 nm) with HWG coupling. Green and yellow regions indicate corresponding pockets of intramuscular fat between the photograph and PA images. The dotted white line indicates the surface of the sample and boundary with coupling medium. Each image is scaled to its own maximum value above the 0 dB noise floor. (c) Average PA spectrum of encircled regions of intramuscular fat using HWG as coupling medium with added fluence correction and noise floor indicated by black dotted line.