Automatic detection of malignant cells in soft tissues from microscopic images

The automatic detection of malignant cells in soft tissues with the help of medical images assists the pathologists in histopathology diagnosis to confirm cancer. This system helps pathologists to improve the accuracy, efficiency to detect malignancy and to minimize the inter-observer variation. In addition, the method may help us to analyze the morphometric and staining characteristics of the cells. The most reliable screening method, histological microscopic images from a biopsy sample is stained, based on the immunohistologically stained slide are fed to the microscope then images is acquired. The statistical texture features are extracted from microscopic images. The extracted features are fed to the radial basis function (RBF) network for classification and particularly suitable for use in an expert system that aids in the diagnosis malignancy of microscopic images. The overall accuracy of classification of the proposed approach is 83.3%.In this work, effective method of automatic detection of malignant cells in soft tissues is developed with the help of microscopic images. The advantages of this work are the time taken for detection is minimized and an early treatment for the disease is achievable.