Prostate Cancer Classification Using Artificial Neural Networks

Abstract

Prostate cancer is one of the major causes of cancer death in men in the Western world. One in 6 men will develop prostate cancer in his lifetime. Early detection and accurate diagnosis of the disease can improve cancer survival and reduce treatment costs [Computer-aided detection of prostate cancer in MRI. IEEE, May 2014]. The artificial neural network (ANN) is used to classify and extract the feature of the magnetic resonance imaging (MRI) scan. This classification is computer-based and has the advantage that it enriches the patient from the doctor and undergoes the different diagnostic techniques that may appear to be stressful and painful. A computer-aided classification set of MRI scans of the prostate subjects these images to a set of preprocessing and image enhancement giving an accurate classification to improve the properties of the image such as noise filtering to improve the contrast of the image and then using segmentation to separate the tumor region and feature extraction; this step can be done by using a set of rolls (explained below). This feature is used to input data to ANN. It classifies the given data set into benign or malignant (prostate cancer). The system succeeded in classifying malignant tumors with 92.4%, increasing decision-making and providing particular treatment. The computerized method detects the tumor in its early stage; thus, good classification is an essential part. The first stage is to extract features from MRI using principal component analysis (computer aided diagnosis), and the second stage is to train the probabilistic neural network (ANN) for classification. The computer aided diagnosis method gives existing attributes a new set of calculated parameters.