Multispectral Photoacoustic Imaging Protocol Evaluation

To characterize the multispectral PA capabilities of the TriTom, three PAI contrast agents, with distinct optical absorbance spectra, were prepared with concentrations adjusted to match an optical density (OD) peak of $1{\mathrm{cm}}^{-1}$ per sample at their peak optical absorption wavelengths: ${\mathrm{CuSO}}_4{·5\mathrm{H}}_2\mathrm{O}$ (99.2 mM, 810 nm), indocyanine green (ICG) ( $26.5\mu \mathrm{M}$ , 780 nm), ${\mathrm{NiSO}}_4·6{\mathrm{H}}_2\mathrm{O}$ (13.6 M, 720 nm), and a control sample of DI water. The samples’ optical spectra were measured using a spectrophotometer (GENESYS 30, Thermo Fisher Scientific, Waltham, Massachusetts, United States). A phantom was assembled by filling PTFE microcuvettes with each sample and placing them into a sample holder, then mounting the holder into the TriTom. Multispectral PA scans were then acquired from 690 to 890 nm at intervals of 10 nm with an average laser energy of 80 mJ per pulse for each wavelength scan.
A PA spectrum for each sample was extracted from the reconstructed volumes of the multispectral phantom. Using the 800-nm volume, the sample microcuvettes’ areas were identified and segmented, from which the mean target inclusion of each sample across all wavelength volumes was calculated. The mean target inclusions were background subtracted to produce PA spectra for the samples. The resulting PA spectra were normalized to the ${\mathrm{CuSO}}_4{·5\mathrm{H}}_2\mathrm{O}$ spectrophotometer data to evaluate the PA spectral accuracy of the system without considering the reconstruction algorithm’s lack of physical corrections.