MR Image Analysis in MS: Identification of a Surrogate

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Multiple sclerosis (MS) is the most common demyelinating disease in humans and has a complex clinical course that includes unpredictable relapses and variable remissions. This makes clinical evaluation of MS difficult. Therefore, current clinical trial designs must incorporate large numbers of patients followed over long periods. These designs are expensive and may deprive patients of timely access to effective treatment. The use of robust surrogate marker(s) that have predictive value could reduce problems in evaluating new drugs and improve the management of individual patients. MRI-based measures such as volumes of lesions, black holes, contrast enhancements, atrophy, and magnetization transfer ratios, are expected to serve as robust surrogates. However, a number of studies have shown that the correlation between these MR measures and clinical score is weak. We hypothesize that this weak correlation is in part due to the use of improper image analysis tools necessary for robust image quantitation and in part due to failure to define the correct MRI surrogate. In these studies we propose to develop an integrated image analysis package that is robust and automatic for accurate quantitation of tissue volumes. An important feature of this analysis package is its ability to analyze images acquired on a wide range of MR scanners using a plethora of MR sequences, greatly extending its utility. This package allows us to follow temporal changes in individual lesions, as well as currently used global changes. This analysis package will be rigorously evaluated using an extensive database that contains images on more than 2,000 MS patients, followed over several years. Using this database, we propose to identify surrogate(s) based on individual or some combination of MRI-measures. Finally, this software will be distributed to a few select centers for multicenter evaluation. While the main emphasis is on MS, this system should be readily adaptable to investigate and manage various neurological disorders that require accurate determination of tissue volumes and their temporal change.