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.

中文翻译：

---

乳腺X光片结构相似性的性能分析

本文介绍了一种在数字乳房X线照片中发现微钙化的新方法，其中使用看不见的上手变换和合成神经网络的混合来设计分类器。微钙化诊断实际上是通过从图形曲线系数中绘制出微钙化房屋来进行的，并且这些结果实际上被用作语义网络输入以进行区分。该神经网络具有一个输入，2个隐藏层以及一个结果。人体将乳房X线照片的图像归类为健康或不常见，不规则强度归类为可治愈或致命。实验表明，我们的技术可以轻松提供更好的结果。该系统实际上是在乳腺X射线摄影图像评估数据源中检查的。