Research On Image Processing Problems Of Biomedical Photoacoustic Tomography

In recent years, photoacoustic imaging as a new medical imaging modality has developed gradually. Because of photoacoustic effect of biological tissue imaging, it is non-invasive, non-ionizing radiation, and can be performed easily. It primarily reflects distribution of light absorption in biological tissues that effectively provides important information in biological tissue like oxygen saturation concentration and hemoglobin content, etc. It can also perform functional imaging, which has great prospects for development in medical diagnosis and treatment. Currently, photoacoustic imaging has been used in the field of brain imaging, breast imaging, joint diseases diagnose lymphatic disease diagnosis and endoscopic technologies emerge, showing great vitality. Especially in the past twenty years, photoacoustic imaging has aroused great attention in many domestic and foreign research institutions, and the correspondingresearch results continue to emerge. It can be said that for photoacoustic imaging, it is the prime time of development now.In photoacoustic imaging, image quality is the key factor affecting its development. In many imaging algorithms, the problem of determining speed of sound in heterogenous tissue are not well solved, affecting the image reconstruction. In order to solve such a problem, we propose two methods to optimize photoacoustic image. In this paper, we first propose a fitted delay compensation based optimization method, then a multi-sampled based optimization method is also developed. The former, from the perspective of the time domain, is a linear delay compensation based method, which focuses on the compensation of acoustic propagation time to make up for the delay time caused by uneven sound velocity calculation error. The latter, from the perspective of the signal processing, process the signals which are repeatedly sampled, to reduce the intensity noise by aligning interpolation between multi frames. In order to validate the corresponding optimization algorithm, this paper designed several experiments to verify. Firstly we validate it by using the specific phantom model. Further, we made verification by taking experiment on human finger joint. Based on the actual experimental results, it can be shown that the optimization method proposed in this paper have a significant effect in image optimization in acoustically inhomogeneous media.Further, photoacoustic and ultrasound hybrid imaging will be a new imaging modality to be realized in the future, so we investigate it in this paper. Based on Verasonics platform, we explored two imaging ways to implement the hybrid imaging. One is carried out on a single host computer. In this case, photoacoustic and ultrasound reconstruction run on the same host computer, and are also displayed on the same machine. The other is a splited hybrid imaging, in which case, photoacoustic and ultrasound reconstruction are carried out on different hosts. We call them comprehensive and separate hybrid imaging repectively. We designed specific experiments and test image quality as well as frame rate. Corresponding experimental results show that the two methods could both effectively perform hybrid imaging, showing pretty good result.