DIAGNOSIS OF NORMAL PRESSURE HYDROCEPHALUS BY AUTOMATED PROCESSING OF MRI IMAGES

摘要

Discussed herein is a parametric model for DTI MD histogram fitting, named the Generalized Voss-Dyke function, which is highly successful in segregating NPH cases from potential confounders without reliance on operator dependent region-of-interest analyses or inter-subject registration. The Generalized Voss-Dyke function is useful for managing the imaging of any tissue interfaces.

主权项

1. A method to differentiate NPH from related conditions comprising:
generating, by a computing device, a histogram of voxels in an image from a scan of a patient's brain; fitting, by the computing device, a curve to the histogram of voxels using a model of brain characteristics, wherein the model comprises at least three functions expressing distribution-groups for voxels and a weighting variable for each of the at least three functions, wherein the at least three functions include a function representing brain tissue, a function representing cerebrospinal fluid (CSF), and a function representing a mix of brain tissue and CSF, and wherein the function representing the mix of brain tissue and CSF enables unequal partial volume voxel distributions of the brain tissue and the CSF; iteratively changing, by the computing device, the function representing the mix of brain tissue and CSF and a variable within the function representing the mix of brain tissue and CSF that represents a relative amount of brain tissue versus CSF mixed in the voxels such that a best fit to the histogram curve is obtained; and comparing, by the computing device, at least the weighting variable for the function representing the mix of brain tissue and CSF to pre-determined values representative of one or more brain conditions or disorders.