Tissue type characterization using photoacoustic power spectrum, a feasibility study

The development of technologies capable of non-invasive characterization of tissue has the potential to significantly improve diagnostic and therapeutic medical interventions. In this study we investigated the feasibility of a noninvasive photoacoustic (PA) approach for characterizing biological tissues. The measurement was performed in the transmission mode with a wideband hydrophone while a tunable Q-switched Nd:YAG pulsed laser was used for illumination. The power spectrum of photoacoustic signal induced by a pulsed laser light from tissue was analyzed and features were extracted to study their correlation with tissue biomechanical properties. For a controlled study, tissue mimicking gelatin phantoms with different densities and equivalent optical absorptions were used as targets. The correlation between gelatin concentration of such phantoms and their mechanical properties were validated independently with a dynamic mechanical analyzer capable of calculating complex loss and storage moduli between two compression plates. It was shown that PA spectrums were shifted towards higher frequencies by increasing gelatin concentration. In order to quantify this effect, signal energy in two intervals of low and high frequency ranges were calculated. Gelatin concentration was correlated with PA energy in high frequency range with R²=0.94. Subsequently, PA signals generated from freshly resected human thyroid specimens were measured and analyzed in a similar fashion. We found that in aggregate, malignant thyroid tissue contains approximately 1.6 times lower energy in the high frequency range in comparison to normal thyroid tissue (p<0.01). This ratio increased with increasing illumination wavelength from 700 nm to 900nm. In summary, this study demonstrated the feasibility of using photoacoustic technique for characterizing tissue on the basis of viscoelastic properties of the tissue.