Performance Analysis of Structural Similarity in Mammograms

Abstract

This article introduces a new method to microcalcification discovery in digital mammograms, in which the classifiers are designed using the blend of unseen upper hand transformation and synthetic neural networks. Microcalcification diagnosis is actually carried out through drawing out the microcalcification homes from the graphic curve coefficients, as well as these results are actually used as semantic network input for distinction. The neural network has one input, 2 hidden layers as well as one outcome. The body classifies mammography graphics as healthy or uncommon and also the irregular intensity as curable or fatal. Experiments reveal that our technique can easily deliver a much better result. The system is actually examined in the Mammography Image Evaluation data source.