Fuzzy clustering of CT images for the measurement of hydrocephalus associated with tuberculous meningitis

摘要

Computed tomography is used as an aid in the diagnosis of tuberculous meningitis (TBM) for the examination of a number of visual indicators of the disease. We present an algorithm that uses modified fuzzy c-means clustering to segment CT images of the brain into different tissue types. The ventricle/brain ratio is then calculated to measure hydrocephalus, a common sign of TBM in children.