Detection of Hydrocephalus Lateral Ventricles Quantitatively in Brain MRI images of Infants

In this paper, a new clustering method is applied to segmentation and volume change assessment in lateral ventricles affected by hydrocephalus pathology in the infants. The hydrocephalus is characterized as such, which results from excessive accumulation of cerebrospinal fluid in the ventricles, leading to enlargement of cerebral ventricles. The complex shape o the ventricular system is evaluated by a visual assessment of MRI scans, but still there is necessity to note the amount of change in volume. We present a algorithm with adjustable feature weight with Optimal feature selection using validity function. The algorithm is called Optomal feature Weight adjustable FCM with Gaussian smoothing (OfwaFCM). There was need to develop new algorithm for segmenting non-symmetric images. The objective of this algorithm is to produce fine clustering results and to reduce effect of noise. In this method, clustering centroids and functions are used to evaluate the clustering results and validity fuction based on fuzzy partitioning is used to select Optimal Feature Weights. The results produced states that the proposed scheme provides Volume calculation and clustering performances for Detecting and segmenting the hydrocephalus.