Image Processing Reference
In-Depth Information
outcomes: cancer or non-cancer.
Algorithms based on fuzzy image processing are first applied to enhance the contrast of the
original image. Then the features characterizing the underlying texture of the interesting
regions are built. Feature extraction is derived from the gray-level co-occurrence matrix.
Rough set approach for attribute selection and rule generation is used. Rough neural net-
work is designed to discriminate different regions of interest and test whether they are cancer
or non-cancer specific. The application of breast cancer imaging is chosen as a case study.
Finally, a rough neural network is designed to discriminate different regions of interest in
order to separate them into malignant and benign cases. Experimental results show that
the introduced scheme is very successful and has high detection and classification accuracy.
The results show that the rough hybrid approaches have high detection accuracy, reaching
over 98%.
Search WWH ::




Custom Search