Paraconsistent Extractor of Mammographic Images Applied in the Process of Diagnosis of Breast Cancer Assisted by Computer

In this expository work, we show an application of a new class of ANN, namely the Paraconsistent Artificial Neural Network - PANN. Also, we use an algorithm - the Paraconsistent Extractor - for our studies. It was performed on the attributes of mammographic images. To perform these simulations, we used two different databases. The first one is used to classify calcifications, is composed of 143 samples divided into 64 benign cases and 79 malignant cases represented by form. The second is intended for mammographic masses and tumors classification and is composed of 57 regions of interest divided into 37 malignant and 20 benign cases, represented by form factors, transition edges and texture measures. The results demonstrate the qualities of Paraconsistent classifier when using a small number of samples for training the neural network and its low processing time. The proposed classifier can be rated as a Computer-Aided Diagnosis system (CAD). The Paraconsistent Extractor can obtain image parameters and sends them to the paraconsistent artificial neural network to analyze them.