Fuzzy Neural Networks for Diagnosis of Malignant MesotheliomaIntelligent Engineering Systems Through Artificial Neural Networks Conference (2007)
Many computer based diagnostic systems are being used in practice in areas such as the mammography, analysis of magnetic resonance imaging (MRI), and computer tomography. However, the diagnosis of pathological images is still more of an art, and it is based on the pathologist's knowledge, experience, and intuition. In pathology, suspect tissue samples are removed from the host and are stained before they are observed under the microscope. In this paper, we propose a computer based diagnostic system to analyze images of stained samples. The proposed system tries to mimic the human vision and decision making process. During preprocessing, pixels are mapped from the red-blue-green (RGB) color space to the hue-saturation-intensity (HIS) feature space. The feature extraction stage deals with obtaining color features from the segmented images. During the decision making, the images are separated in two classes positive and negatives.
- fuzzy neural nets,
- patient diagnosis
Publication DateNovember, 2007
LocationSt. Louis, MO
Citation InformationArun D Kulkarni and Madhukar Bandi. "Fuzzy Neural Networks for Diagnosis of Malignant Mesothelioma" Intelligent Engineering Systems Through Artificial Neural Networks Conference (2007)
Available at: http://works.bepress.com/arun-kulkarni/48/