Telemedicine system of pathology image using the Internet

In this study, tumor images are classified into 4 groups according to the symptom automatically in order to construct the telemedicine system of pathology image using the Internet. As this method, we use the texture analysis for obtaining parameters and the discriminant analysis for the automatic classification. The discriminant analysis based on 5 parameters for 103 tumor images resulted in the correct automatic classification of 81.5%. From the result of this study, it has provided the possibility of the image-based diagnosis of tumor with texture analysis.