| Paper Details: | Downloads: 563 |
| Serial Number: | P1160934843
|
| Title: | Development of a new composite feature vector for the detection of pathological and healthy tissues in FLAIR MR images of brain
|
| Authors: | Nandita Pradhan and A.K. Sinha
|
| Abstract: | This paper presents a new technique for segmentation and detection of pathological tissues (Tumor and Edema), normal tissues (White Matter and Gray Matter) and fluid (Cerebrospinal Fluid) from Fluid Attenuated Inversion Recovery (FLAIR) magnetic resonance (MR) images of brain with the help of new composite feature vectors comprising of wavelet and statistical parameters, in contrast to other researchers who developed feature vectors either using statistical parameter or using wavelet parameters. The main contributions of paper are (1) Segmentation of intracranial brain image in five segments is done with k-mean algorithm which is based on combined features of wavelet energy function & statistical parameters and are very useful in describing texture properties and pixel value analysis of different segments of images, giving better result. (2) In addition to tumor, edema is also characterized as a separate class which is critical for therapy planning, surgery, diagnosis and treatment of tumors. Block processing of image is done by extracting feature vectors from small blocks of 4×4 pixels of image corresponding to tumor, edema, white matter, gray matter and cerebrospinal fluid and training artificial neural network using back propagation algorithm. The result is very satisfactory with minimum error coming to be as low as 4.36e-017.
|
| Keywords: | FLAIR Magnetic Resonance Image, Segmentation, k-means algorithm, Edema, Wavelet Energy function
|
| Journal/Conference: | International Journal of Bioinformatics and Medical Engineering
|
| Volume: | 10
|
| Issue: | 1
|
| Submission Date: | 8/19/2009 12:00:00 AM
|
| Review Date: | 11/26/2009 12:00:00 AM
|
| Publishing Date: | 2/8/2010 12:00:00 AM
|
| Article Downloads: | 563
|
| Download: |
|