Photoacoustic Imaging In Ultrasonic Signal Acquisition And Preprocessing Technology Research

Photoacoustic tomography (PAT) has become a novel noninvasive medical imaging technology. It combines the advantages of pure optic imaging and pure ultrasound imaging and can provide tissue image of high resolution and high contrast tissue. The key technologies in photoacoustic tomography are signal acquisition and image reconstruction. Based on the aspects of signal acquisition and data preprocess, this research contents in this thesis are as follows:At first, a photoacoustic tomography system of high accuracy based on the architecture of "Unit Rotation Scanning" has been designed and implemented in this thesis. The system is a platform composed of digital oscilloscope, tunable laser and rotary table which are programmable. And the controlling software has also been developed. The problems of ultrasonic sensor’s cable shielding and response delay between devices have been solved. The advantages are saving the cost of hardware and making the system coordination run better. This method of constructing the photoacoustic tomography system is expected to be a noninvasive biological tissue and functional imaging system.In the process of photoacoustic tomography on the above platform, we found that the reconstructed image had artifacts when the water was shallow. The mechanism of artifacts in the photoacoustic tomography is analyzed based on the propagation characteristics of ultrasonic waves in the water. The analysis shows that artifacts are generated from the delay caused by the reflection of the ultrasonic signal upon the water surface. Then a method through the delay subtraction is proposed to control these artifacts. The reference is also provided on how to control the water height in the practical application. The method is validated with our experimental platform. From the experiment, it is shown that the proposed algorithm can eliminate artifacts effectively and the quality of the reconstructed image is better.At last, to overcome the problem of the noise wave in the ultrasound data collected from the photoacoustic tomography system, a new method is proposed to firstly separate ultrasound signals and noise wave based on Renyi entropy and then to reconstruct the image using the separated signals. The experiment on the photoacoustic tomography platform shows that the noisy part in the ultrasound data acquisition can be segmented out through the threshold obtained by the histogram of Renyi entropy. The method is validated through image quality assessment. It is proved that the proposed algorithm can segment ultrasound signals effectively and enhance the quality of the reconstructed image.