The characteristic microstructures in biological tissues could be used to differentiate tissue types, such as tumor vs.
normal tissue. The spatial resolution of classical photoacoustic tomography (PAT) mainly depends on the wavelengths of
the detected ultrasonic signals. In order to present the very detailed microstructures in a biological sample, the receiving
bandwidth of the PAT system needs to be extremely wide. Another challenge in detecting the high frequency signals
associated with microstructures is the strong acoustic attenuation which increases quadratically with ultrasound
In this study, we propose a novel photoacoustic spectral analysis (PSA) technique which evaluates the microstructures in
tissues by analyzing the spectral parameters of detected photoacoustic signals. Experimental result verified that, using a
limited 1-5 MHz working bandwidth, PSA could effectively differentiate two melanoma-mimicking phantoms
containing different microstructures (49 μm and 199 μm absorber sizes respectively). In comparison, since the physical
scales of the microstructures are too small and beyond the spatial resolution of the PAT system, classical tomographic
imaging could not differentiate the two phantoms. The findings from this study suggest that the proposed PSA technique
could help distinguish different tissue types, by evaluating the characteristic microstructures in tissues, without relying
on the detection of high frequency signals which is extremely challenging when the target object is deep.