Chronic liver disease is a worldwide health problem, and hepatic fibrosis (HF) is one of the hallmarks of the disease. Pathology diagnosis of HF is based on textural change in the liver as a lobular collagen network that develops within portal triads. The scale of collagen lobules is characteristically on order of 1mm, which close to the resolution limit of in vivo Gd-enhanced MRI. In this work the methods to collect training and testing images for a Hotelling observer are covered. An observer based on local texture analysis is trained and tested using wet-tissue phantoms. The technique is used to optimize the MRI sequence based on task performance. The final method developed is a two stage model observer to classify fibrotic and healthy tissue in both phantoms and in vivo MRI images. The first stage observer tests for the presence of local texture. Test statistics from the first observer are used to train the second stage observer to globally sample the local observer results. A decision of the disease class is made for an entire MRI image slice using test statistics collected from the second observer. The techniques are tested on wet-tissue phantoms and in vivo clinical patient data.

Chronic liver disease is a worldwide health problem, and hepatic fibrosis (HF) is one of the hallmarks of the disease. The current reference standard for diagnosing HF is biopsy followed by pathologist examination; however, this is limited by sampling error and carries a risk of complications. Pathology diagnosis of HF is based on textural change in the liver as a lobular collagen network that develops within portal triads. The scale of collagen lobules is characteristically in the order of 1 to 5 mm, which approximates the resolution limit of in vivo gadolinium-enhanced magnetic resonance imaging in the delayed phase. We use MRI of formalin-fixed human ex vivo liver samples as phantoms that mimic the textural contrast of in vivo Gd-MRI. We have developed a local texture analysis that is applied to phantom images, and the results are used to train model observers to detect HF. The performance of the observer is assessed with the area-under-the-receiver-operator-characteristic curve (AUROC) as the figure-of-merit. To optimize the MRI pulse sequence, phantoms were scanned with multiple times at a range of flip angles. The flip angle that was associated with the highest AUROC was chosen as optimal for the task of detecting HF. (C) The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

The statistical evaluation of molecular imaging approaches for detecting, diagnosing, and monitoring molecular response to treatment are required prior to their adoption. The assessment of fluorescence-enhanced optical imaging is particularly challenging since neither instrument nor agent has been established. Small animal imaging does not address the depth of penetration issues adequately and the risk of administering molecular optical imaging agents into patients remains unknown. Herein, we focus upon the development of a framework for OAIQ which includes a lumpy-object model to simulate natural anatomical tissue structure as well as the non-specific distribution of fluorescent contrast agents. This work is required for adoption of fluorescence-enhanced optical imaging in the clinic. Herein, the imaging system is simulated by the diffusion approximation of the time-dependent radiative transfer equation, which describes near infra-red light propagation through clinically relevant volumes. We predict the time-dependent light propagation within a 200 cc breast interrogated with 25 points of excitation illumination and 128 points of fluorescent light collection. We simulate the fluorescence generation from Cardio-Green at tissue target concentrations of 1, 0.5, and 0.25 µM with backgrounds containing 0.01 µM. The fluorescence boundary measurements for 1 cc spherical targets simulated within lumpy backgrounds of (i) endogenous optical properties (absorption and scattering), as well as (ii) exogenous fluorophore crosssection are generated with lump strength varying up to 100% of the average background. The imaging data are then used to validate a PMBF/CONTN tomographic reconstruction algorithm. Our results show that the image recovery is sensitive to the heterogeneous background structures. Further analysis on the imaging data by a Hotelling observer affirms that the detection capability of the imaging system is adversely affected by the presence of heterogeneous background structures. The above issue is also addressed using the human-observer studies wherein multiple cases of randomly located targets superimposed on random heterogeneous backgrounds are used in a “double-blind” situation. The results of this study show consistency with the outcome of above mentioned analyses. Finally, the Hotelling observer’s analysis is used to demonstrate (i) the inverse correlation between detectability and target depth, and (ii) the plateauing of detectability with improved excitation light rejection.