The goal of this R21 is to obtain sufficient measurements of MR properties of the cirrhotic liver in various states of disease to obtain an estimate of the parameters to develop the experimental design for an R01 focused on early detection of hepatocellular carcinoma (HCC). The statistical properties of a variety of MR parameter will be measured for cirrhotic tissue, premalignant nodules and HCC nodules. The orthogonality of each measure will be assessed. The significance of this problem arises from the prevalence and aggressiveness of HCC which is the most common intraadominal tumor in the world. When detected clinically the disease is deadly, having a 5 year survival rate of only 5%. However, if detected early HCCs can be effectively treated with liver transplantation or tumor resection. HCCs usually arise as the end stage of a pathological process in the cirrhotic liver. Being able to detect dysplastic nodules (premalignant) and early stage HCCs would allow timely and effective therapy for patients. Unfortunately, early radiological detection of HCC is challenging. The primary radiology techniques for detecting HCCs rely on tumor hypervascularity. However, dysplastic nodules and many small HCC nodules do not demonstrate angiographic hypervascularity. Therefore, alternative mechanisms must be exploited for early radiological detection of these cirrhotic nodules. Our hypothesis is that the detection of premalignant and early stage HCCs can be facilitated with a multispectral magnetic resonance imaging (MR1) model. To achieve this end we use rapid imaging techniques to quantify the dynamic T1 contrast enhancement and endogenous properties (T1, T2, T2*, proton density, diffusion, and fat content) of cirrhotic tissue and nodules. We quantify the range and variability of observed values for each of these characteristics as well as the repeatability of these measurements. We examine whether a subset of these MR feature can be used in a multispectral model for improved imaging of HCCs. Using accurate registration of the MR images to each other and to the explanted liver specimen, these radiological measurements will be correlated with detailed pathology observations including both histological and genetic features. We will be determine which MR characteristics are most informative of dysplastic nodules and HCCs will be examined. Comparisons to nodule detection with X-Ray CT will be made. Finally, the relationship between the MR parameters and the underlying histological and genetic features of HCCs, dysplastic nodules and cirrhotic tissue will be assessed.
| Wu, Huadong; Krasinskas, Alyssa M; Tublin, Mitchell E et al. (2005) Registering liver pathological images with prior in vivo CT/MRI data. Med Image Comput Comput Assist Interv 8:564-71 |
| Christina Lee, Wen-Chi; Tublin, Mitchell E; Chapman, Brian E (2005) Registration of MR and CT images of the liver: comparison of voxel similarity and surface based registration algorithms. Comput Methods Programs Biomed 78:101-14 |
