Image-quality assessment in optical tomography

Modern medical imaging systems often rely on complicated hardware and sophisticated algorithms to produce useful digital images. It is essential that the imaging hardware and any reconstruction algorithms used are optimized, enabling radiologists to make the best decisions and quantify a patient's health status. Optimization of the hardware often entails determining the physical design of the system, such as the the locations of detectors in optical tomography or the design of the collimator in SPECT systems. For software or reconstruction algorithm optimization one is often determining the values of regularization parameters or the number of iterations in an iterative algorithm. In this paper, we present an overview of many approaches to measuring task performance as a means to optimize imaging systems and algorithms. Much of the work in this area has taken place in the areas of nuclear-medicine and x-ray imaging. The purpose of this paper is to present some of the task-based measures of image quality that are directly applicable to optical tomography.