A new method for colourizing of multichannel MR images based on real colour of human brain

In this paper, we propose a novel approach for colourizing of multichannel MR images based on real colour of human brain. At first, we use independent component analysis (ICA) to transfer three MR modalities into different representations, which enhance the physical characteristics of tissues. Then, Statistical Features are extracted from three independent components and their wavelet coefficients to generate a feature vector for each pixel of images. Huge amount of features are reduced by using some feature reduction methods such as Principle Component Analysis (PCA) to classify the brain into 3 primary classes. The colour space of Visible Human Project (VHP) dataset is quantized to 10 colours by using a partially supervised algorithm included self-organization map and Linear Vector Quantization (LVQ) algorithms. Acquired 10 colours are allotted to each pixel based on texture and intensity of image in two different steps. Consequently, a fast-automated algorithm is proposed, which has low sensitivity to variation of pixel intensity and precise result regarding to VHP images.