W. Guo and A. Watson (Australia)
Medical Image Processing, ModifiedConjugate Directional Filtering (MCDF), Statistics,Supervised Learning
We recently proposed a method named the modified conjugate directional filtering (MCDF) to carry out the enhancement of conjugate linear features. Tests over some synthetic and real examples have shown that MDCF is not only powerful in enhancing conjugated linear features that is the original design goal of MCDF, but also very useful in enhancing non-linear features. In this paper we present the trial results of using MCDF in medical image processing. The medical image used is a chest radiograph. MCDF(add1) operation is used for processing. Given the fact that the authors are not familiar with any radiologic diagnosis, our effort is to increase the intuitive interpretability of the processed images in contrast to the original image. The improvement is mainly measured by means of statistical indicators in terms of redistribution of the grey values in the images. Its implication in radiologic diagnosis is beyond our knowledge, which should leave open for radiologists to reveal.
Important Links:
Go Back
