Fractal analysis: fractal dimension and lacunarity from MR images for differentiating the grades of glioma
No Thumbnail Available
Date
2015-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Physics in medicine and biology
Abstract
Glioma, the heterogeneous tumors originating from glial cells, generally
exhibit varied grades and are difficult to differentiate using conventional
MR imaging techniques. When this differentiation is crucial in the disease
prognosis and treatment, even the advanced MR imaging techniques fail to
provide a higher discriminative power for the differentiation of malignant
tumor from benign ones. A powerful image processing technique applied to
the imaging techniques is expected to provide a better differentiation. The
present study focuses on the fractal analysis of fluid attenuation inversion
recovery (FLAIR) MR images, for the differentiation of glioma. For this, we
have considered the most important parameters of fractal analysis, fractal
dimension and lacunarity. While fractal analysis assesses the malignancy
and complexity of a fractal object, lacunarity gives an indication on the
empty space and the degree of inhomogeneity in the fractal objects. Box
counting method with the preprocessing steps namely binarization, dilation
and outlining was used to obtain the fractal dimension and lacunarity in
glioma. Statistical analysis such as one-way analysis of variance (AVOVA)
and receiver operating characteristic (ROC) curve analysis helped to compare
the mean and to find discriminative sensitivity of the results. It was found
that the lacunarity of low and high grade gliomas vary significantly. ROC
curve analysis between low and high grade glioma for fractal dimension
and lacunarity yielded 70.3% sensitivity and 66.7% specificity and 70.3%
sensitivity and 88.9% specificity respectively. The study observes that fractal dimension and lacunarity increases with increase in the grade of glioma and
lacunarity is helpful in identifying most malignant grades.
Description
Keywords
box counting, fractal analysis, fractal dimension, glioma, lacunarity
Citation
Smitha KA, Gupta AK, Jayasree RS. Fractal analysis: fractal dimension and lacunarity from MR images for differentiating the grades of glioma. Physics in medicine and biology. 2015;60(17):6937- 47