Search

The two blood flow imaging techniques, laser Doppler blood flow imaging (LDI) and laser speckle contrast blood flow imaging (LSCI), are well established and broadly applied in medical research. They are similar as both detect and process a fluctuating interference (speckle) pattem. However, the difference between processing algorithms provides different imaging characteristics. LDI can provide accurate, quantitative blood flow measurement which is seldom achieved by LSCI. Nevertheless, the fast imaging speed and simple instrumental setup provided by LSCI overcome some of the limitations ofLDI. With the development of high frame rate cameras full field LDI is now feasible and with the development of new processing algorithms LSCI is now providing more accurate quantitative information. It is therefore important to compare the performance of these two techniques. A full-field LDI system based on an FPGA (Field Programmable Gate Array) coupled with a high-speed CMOS (Complementary Metal-Oxide-Semiconductor) camera chip has been developed which provides blood flow images with flexible frame rates and spatial resolution. When a high spatial resolution is required, 1280xl024-pixel blood flow images were obtained by processing up to 2048 samples at O.2fps (frame per second). Altematively, a maximum of 15.5fps was achieved by reducing the resolution and sampling points to 256x256 pixels and 128 samples respectively. As a generic full-field LDI system, several parts of the system (memory unit, processing unit) can be simply updated or transplanted to another platform. The resource usage is optimized by utilizing a mixture of fixed and floating-pointing implementations, and the imaging speed is maximised because of the design of streamline structure which enables continuous input of data. Images were obtained of rotating diffusers at different rotation velocities and the system provides a linear relationship with velocity. Human blood flow images are also demonstrated both of the finger and of a healing wound. The author-designed LDI system was then applied to a high-spatial resolution flow imaging application in which the mixture of water and polystyrene micro spheres was pumped through a micropipette (diameter = 250llm) with controlled velocities, and the resulting flow was imaged and processed. The accurate, high-spatial resolution flow measurement was demonstrated by the resulting flow images which are of size 1280x 1 024 pixels and obtained by processing 2048 samples at each pixel. Besides the LDI system, a novel LSCI system has been developed on the same platform, establishing a unique LDI and LSCI hybrid system. By developing the LSCI method with equivalent exposures, the LDI data can be analysed using LSCl processing, enabling a truly fair comparison of these two methodologies. For comparison, measurements were carried out on a rotating diffuser that simulates the human tissue with controlled parameters. Although LDI and LSCI are qualitatively similar, the lack of quantitative blood flow measurement ofLSCI was recognized from the comparison since LSCI is exposure time dependent and unable to linearly detect the velocity changes. 11 To improve the linearity and accuracy ofLSCI measurement, multi-exposure laser speckle contrast imaging (MLSCI) has been introduced. However this increases image acquisition time as consecutive images at different exposure times need to be acquired. On the basis of the novel LSCI method, a new MLSCI scheme has been invented. The advantage of the MLSCI is that each frame is exposed with a fixed duration and various exposure times are alternatively achieved by accumulating several successive frames. In this way, the requirement to obtain a wide range of exposure times from consecutive images is overcome. This reduces image acquisition time as it depends on the longest exposure time rather than the sum of all exposures. From measurements of a rotating diffuser, the MLSCI was demonstrated to be capable of quantitatively measuring flow changes as in LDI. III