QUANTITATING DISEASE PROGRESSION FROM THE MRI IMAGES OF MULTIPLE SCLEROSIS PATIENTS

摘要

Methods and systems are provided for the automated detection and analysis of structural tissue alterations related to myelin and axons/neurons in one or more biological structures of a patient's nervous system obtained from data from a medical imaging system, or the initial sensing or data collection processes such as, those that could be used to generate an image. In some embodiments, the method comprises, at a system having a memory and one or more processor for processing and displaying images of the biological structure, computationally processing at least a T1 weighted magnetic resonance image of the structure and a T2 weighted magnetic resonance image of the structure in order to analyze at least a portion of the structure of the nervous system using a plurality of stored tissue classifier elements to determine if the portion of the structure correlates with the presence of myelin. Such methods are useful for the detection of diseases associated with demyelination.

主权项

1. A method of classifying myelin components to determine whether a biological structure of the nervous system contains myelin, the method comprising, at a system having a memory and one or more processor for processing and displaying images of the biological structure: computationally processing at least a T1 weighted magnetic resonance image of the structure and a T2 weighted magnetic resonance image of the structure in order to analyze at least a portion of the structure of the nervous system using a plurality of stored tissue classifier elements developed using statistical modeling to determine if the portion of the structure correlates with the presence of myelin, in which event the portion of the structure is determined to contain myelin, wherein the computational processing includes:
registering the T2 weighted magnetic resonance image with the T1 weighted magnetic resonance image by aligning a voxel in the T2 weighted magnetic resonance image with a corresponding voxel in the T1 weighted magnetic resonance image; and dividing, for each voxel in the registered T2 weighted image, a first intensity value in the registered T2 weighted image by a second intensity value of the corresponding voxel in the T1 weighted magnetic resonance image to generate a map of the biological structure.