In traditional photoacoustic tomography, external illumination is used to excite acoustic waves. However, with the assistance of fibre-transmitted light, multidirectional illumination or internal illumination can be achieved which can obtain a better image at a deeper depth. Laser pulses delivered by fibre are energy-limited by the fibre core size and damage threshold. To increase the amplitude of photoacoustic waves and their penetration, it is necessary to improve the fibre coupling energy and efficiency. To improve the coupling performance of single fibres, we use a cylindrical lens array to homogenize the incident beam before a coupling lens. Simulation in Zemax shows that this approach flattens the beam profile on the front surface of the fibre, decreasing the risk of fibre damage. Experimental results with fibre core diameters of 1mm and 1.5mm show that both types of fibre can output more than 50mJ energy per pulse at 700nm wavelength. The coupling efficiency is measured to be above 70% and even reaches 90% as the wavelength changes from 675nm to 900nm. This improvement of coupling energy in single fibres will benefit photoacoustic tomography applications using internal illumination.
Imaging complicated structures with photoacoustic (PA) modality when the field of view is limited can result in significant imaging artifacts or missing structures. Approaches to solve this problem include new reconstruction algorithms and specific transducer structures, such as hemi-spherical transducer arrays for breast cancer detection. However, most existing PA imaging techniques require either fullview complete projection data collection or complex and computationally-intensive reconstructions. Such approaches are not only time-consuming but also unsuitable for many clinical applications, because most clinical imaging hardware is constrained to limited reconstruction angles. In this paper, we present a method of using two commercial linear array transducers at different orientations to increase the view angle and thus improve the reconstruction of PA imaging. The method involves a two-step process. First, a calibration phantom is imaged to calibrate the relative position of these two linear transducers. Second, two PA images are obtained by a simple back projection algorithm and these images are registered using the information from the calibration process. The final registered image contains more detailed structures without the requirements of a specialized transducer or long processing time. Experimental results show that this method has the potential to provide good image quality using standard low-cost transducers.