Analysis of an Optical Coherence Imaging Modality on the Detection of an Abnormality in Biological Tissue with a Nanoparticle Contrast Agent

Abstract

There is great interest in promoting the use of contrast agents in optical imaging for better diagnosis of diseases. However, until recently, there was still no quantitative method existing to assess the ability of contrast agents in improving clinical diagnosis. In this study, we used the method of task-based medical image analysis as a quantitative tool to evaluate the effectiveness of nanoparticles as contrast agents in an OCT imaging modality for clinical diagnosis.The task was formulated as the detection of abnormalities in a biological tissue using a quadratic observer. We derived the test statistics of the quadratic observer, and an analytical expression for the index of detectability for such a quadratic observer. The statistical properties of the OCT data are determined by the stochastic mechanisms in the imaging system as well as the demodulation method in the data acquisition process. In this analysis, we have considered the effect of phase fluctuations from the broadband source, the shot noise fluctuations of the imaging system, and the scattering noise due to refractive index fluctuation in the biological tissue. Our analysis was performed at the system level by integrating to the analysis the data demodulation process based on a mixer scheme. Also, we implemented the dynamic focusing in the scanning process.Optical propagation in biological samples is dominated by scattering due to fluctuations in refractive index. For OCT imaging, it is assumed that only the singly scattered field from the sample will contribute significantly to the interferometric optical power. We modeled the normal biological tissue (the background) as a spatial Poisson field of randomly distributed scattering centers, and the abnormality (the target) as a region with a different concentration of scattering centers embedded in the background.We presented the results on the detectability of abnormalities of different sizes, with or without the presence of contrast agents. We have shown that the application of nanoparticle contrast agents improved the detectability of small abnormalities which are usually difficult to detect. We have thus shown the efficacy of the task-based analysis framework in delivering quantitative assessments of the efficiency of contrast agents.