The last decade has seen major advances towards clinically relevant optical tomographic imaging methods. For example, it is well established that he most accurate image reconstruction schemes are based on the equation of radiative transfer (ERT). However, the accuracy comes at the cost of long computation times, which have hampered its practical utility. Besides a general overview on how to implement ERT-based reconstruction algorithms, we focus on various method that accelerated the image reconstruction process. PDE-constrained algorithms, SPN methods and parametric reconstruction techniques, as well as domain decomposition and parallel programming approaches will be discussed. All of these approaches combined have reduced the image reconstruction times from hours or even days to minutes or seconds. Furthermore, instrumentation that incorporates digital-signal-processing (DSP) chip technology has significantly increased the signal-to-noise ratios, and ever smaller signals can be detected in shorter times. This opens the doors to novel applications in dynamic optical tomography. Practical examples encountered in clinical and preclinical imaging such as monitoring of tumor growth and regression, effects of anti-angiogenic drugs, breast cancer screening, and detection of arthritis will be presented.
Back to Mathematical Problems, Models and Methods in Biomedical Imaging