Abstract
This paper presents a local feature vector based method for automated Multiple Sclerosis (MS) lesion segmentation of multi spectral MRI data. Twenty datasets from MS patients with FLAIR, T1,T2, MD and FA data with expert annotations are available as training set from the MICCAI 2008 challenge on MS, and 24 test datasets. Our local feature vector method contains neighbourhood voxel intensities, histogram and MS probability atlas information. Principal Component Analysis(PCA) \cite{PCA} with log-likelihood ratio is used to classify each voxel. MRI suffers from intensity inhomogenities. We try to correct this ''bias field'' with 3 methods: a genetic algorithm, edge preserving filtering and atlas based correction. A large observer variability exist between expert classifications, but the similarity scores between model and expert classifications are often lower. Our model gives the best classification results with raw data, because bias correction gives artifacts at the edges and flatten large MS lesions.
Keywords
Source Code and Data
No source code files available for this publication.
Reviews
Simon Warfield
Friday 25 July 2008
This paper describes an approach to the detection of MS lesions from brain MRI.
It utilizes a local feature estimation approach.
Interestingly, the paper finds improved results without the use of intensity inhomogeneity artifact compensation, due to edge artifacts created by the intensity inhomogeneity compensation algorithm.
The results are promising and with further work may be comparable to those of human raters.
Typo:
methode should be method
Martin Styner
Monday 28 July 2008
The paper presents a MS lesion segmentation using a PCA based local feature vector classification incorporating a given voxel neighborhood.
The evaluations show that the method surprisingly performs better without intensity inhomogeneity correction. This may be an artifact of the employed bias correction methods. A tissue classification based bias correction (Well, Leemput, and many others) may help in this case.
The paper could use another revision and feels in part like a technical report.
Minor:
Page 5 top: "affine fitted" should be "affinely fitted"
Page 7 middle: "non-rigid fit" should be "non-rigidly fit"
Conclusion section: "by by" should be simply "by"