Study Methods Input Purpose Dataset Classifier Results Dheeba et al. [3 ] Particle Swarm Optimized Wavelet Neural Network (PSOWNN) Mammogram Improve classification accuracy in breast cancer detection and reducing misclassification rate 216 mammograms PSOWNN (i) Sensitivity 94.167% (ii) Specificity 92.105% (iii) AUC 0.96853 (iv) Youden’s index 0.86272 (v) Misclassification rate 0.063291 Xu et al. [10 ] New algorithm based on two ANNs Mammogram Classification of masses 30 cases and 60 mammograms (containing 78 masses) ANFIS and MLP (i) True positive (TP) rate 93.6% (73/78), (ii) Number of the FPs per image 0.63 (38/60). Alayliogh and Aghdasi [11 ] ANN and biorthogonal spline wavelet Mammogram Classification of microcalcification cluster (MCC) and image enhancement 40 digitized mammogram ANN (i) Sensitivity 93%, (ii) FP rate (MCC/image) 0.82 Dhawan et al. [13 ] (i) ANN (ii) second-order gray-level statistics Mammogram Classification of significant and benign microcalcifications 5 image structure features (i) Three-layer perceptron based ANN The entropy feature has significant discriminating power for classification Chitre et al. [15 ] ANN Mammogram Classification of microcalcification into benign and malignant (i) 40, 60, and 80 training cases (ii) 151, 131, and 111 test cases ANN Neural network is a robust classifier of a combination of image structure and binary features into benign and malignant Kevin et al. [16 ] ANN Mammogram Classification of microcalcifications and nonmicrocalcifications 24 mammograms with each containing at least one cluster of microcalcifications Cascade correlation ANN (CC ANN) (i) TP detection rate for individual microcalcifications is 73% and 92% for nonmicrocalcifications Zheng et al. [18 ] ANN and BBN Mammogram Compare performances of ANN and BBN 3 independent image databases and 38 features ANN and BBN Performance level ( value) (i) ANN value (ii) BBN value 0.845 ± 0.011 (iii) Hybrid classifier (ANN and BBN)A z value increased to 0.859 ± 0.01 Zhang et al. [19 ] Digitize module, detection module, feature extraction module, neural network module, and classification module Mammogram Classification of microcalcification clusters/suspicious areas Fuzzy detection algorithm (i) 30 digital images (15 contain benign cases and 15 contain malignant cases) Backpropagation neural network (BPNN) (i) Fuzzy detection rate (benign 84.10% and 80.30%) (ii) Classification rates (feature vector, is 83.8%), (feature vector is 72.2%) Lashkari [20 ] ANN and Gabor wavelets Mammogram Classification of breast tissues to normal and abnormal classes automatically (i) Images of 50 normal and 50 abnormal breast tissues (ii) 65 cases for training set and 35 cases for testing set ANN and Gabor wavelets (i) Classification rate (testing performance 96.3% and training performance 97.5%) Saini and Vijay [17 ] Image registration technique and ANN Mammogram Classification of benign and malignant 42 mammogram images (30 benign and 12 malignant images) Feed-forward backpropagation and Cascade forward backpropagation artificial neural network Percentage accuracy (feed-forward backpropagation network is 87.5% and Cascade forward backpropagation network is 67.8%) Buller et al. [21 ] Spider web topology with NN Ultrasound Classify and separate benign and malignant lesion 25 sonograms (i) NN classifier (i) 69% accuracy in malignant (ii) 66% accuracy in benign (iii) 66% accuracy in no lesions Ruggierol et al. [22 ] Texture and shape parameter feeds into NN Ultrasound Automated recognition of malignant lesion (i) 41 carcinomas (ii) 41 fibroadenomas (iii) 41 cysts (i) NN classifier (i) 95% accuracy in solid lesions (ii) 92.7% accuracy in liquid lesions Sahiner et al. [23 ] Convolutional NN with spatial and texture image Mammogram Classification of mass and normal breast 168 mammograms (i) Convolution NN classifier (i) Average true positive fraction of 90% at false positive fraction of 31% Chen et al. [24 ] Multilayer feed-forward neural network (MFNN) Ultrasound Classify benign and malignant lesion 140 pathological proved tumors (52 malignant, 88 benign) MFNN (i) 95% accuracy, 98% sensitivity (ii) 93% specificity (iii) 89% positive predictive value (iv) 99% negative predictive value Chen et al. [25 ] Self-organizing map (SOM) Ultrasound Classification of benign and malignant lesions 243 tumors (82 malignant, 161 benign) SOM (i) Accuracy of 85.6, sensitivity 97.6% (ii) Specificity 79.5% (iii) Positive predictive value 70.8% (iv) Negative predictive value 98.5% Chen et al. [27 ] Bootstrap with NN Ultrasound classification of tumor 263 sonographic image solid breast nodules NN (i) Accuracy 87.07%, sensitivity 98.35% (ii) Specificity 79.10% (iii) Positive predictive value 81.46% (iv) Negative predictive value 94.64% Chen et al. [26 ] 2-phase Hierarchical Neural Network (HNN) Ultrasound Differentiate between benign and malignant tumors 1020 images (4 different rectangular regions from the 2 orthogonal planes of each tumor) HNN 4 image analyses of each tumor appear to give more promising result than if they are used separately Chen et al. [28 ] Wavelet transform and neural network Ultrasound Differential diagnosis of breast tumors on sonograms 242 cases (161 benign, 82 malignant) Multilayer perceptron neural network (MLPNN) (i) Receiver operating characteristic (ROC) area index is 0.9396 ± 0.0183 (ii) 98.77% sensitivity, 81.37% specificity (iii) 72.73% positive predictive value (iv) 99.24% negative predictive value Chen et al. [29 ] Multilayer feed-forward neural network (MFNN) Ultrasound Differentiate benign from malignant breast lesions 1st set: 160 lesions 2nd set: 111 lesions MFNN (i) 98.2% training accuracy (ii) 95.5% testing accuracy Joo et al. [30 ] Artificial neural network (ANN) Ultrasound Determining whether a breast nodule is benign or malignant 584 histologically confirmed cases (300 benign, 284 malignant) ANN (i) 100% training accuracy (ii) 91.4% testing set (iii) 92.3% sensitivity, 90.7% specificity Joo et al. [31 ] Digital image processing and artificial neural network Ultrasound Determine breast nodule malignancy 584 histologically confirmed cases (300 benign, 284 malignant) ANN (i) 91.4% accuracy, 92.3% sensitivity (ii) 90.7% specificity Zheng et al. [32 ] Hybrid method (unsupervised k -means cluster, supervised backpropagation neural network (BPNN)) Ultrasound Classification of breast tumors as benign or malignant 125 benign tumors, 110 malignant tumors Combination of k -means with BPNN (i) Recognition rate (94.5% for benign, 93.6% for malignant) (ii) 94% accuracy, 94.5% sensitivity (iii) 93.6% specificity Fok et al. [35 ] ANN with 3D finite element analysis IR Tumor prediction 200 patients ANN Good detection, poor sensitivity Szu et al. [36 ] Unsupervised classification using Lagrange Constraint Neural Network (LCNN) Mid and long IR images Early detection of breast cancer One patient with DCIS LCNN Better sensitivity Jakubowska et al. [37 ] ANN with wavelet transform IR Discrimination of healthy and pathological cases 30 healthy ANN Accuracy (%) (frontal/side) Raw: 90/93, PCA: 90/93 LDA: 93/97, NDA: 93/93 10 with recognized tumors Accuracy (%) (frontal/side) Raw: 80/90, PCA: 80/90 LDA: 90/90, NDA: 80/100 Koay et al. [33 ] Backpropagation NN IR Early detection of breast cancer 19 patients Levenberg-Marquardt (LM) and Resilient Backpropagation (RP) Accuracy (%) (RP/LM) Whole: 95/95 Quadrants: 95/100 Tan et al. [34 ] Fuzzy adaptive learning control network fuzzy neural network IR Early detection of breast cancer and tumor classification 28 healthy, 43 benign tumors, 7 cancer patients FALCON-AART Cancer detection (%) (TH/TDF) Predicted: 95, sensitivity: 100, specificity: 60 Breast tumor detection (%) (TH/TDF) Predicted: 84/71, sensitivity: 33/76, specificity: 91/62 Breast tumor classification (%) (TH/TDF) Predicted: 88/84, sensitivity: 33/33, specificity: 95.5/91 Cardillo et al. [40 ] NN for automatic analysis of image statistics MRI Early detection and classification 150 exams subdivided into 6 groups by contrast NN Better in specificity Tzacheva et al. [41 ] Evaluation of signal intensity and mass properties by NN MRI Automatic diagnosis of tumors 14 patients Feed-forward BPNN 90%–100% sensitivity, 91%–100% specificity, and 91%–100% accuracy Ertas et al. [39 ] Extraction of breast regions by conventional and multistate CNNs MRI Breast density evaluation and abnormality localization 23 women CNN Average precision 99.3 ± 1.8% True positive volume fraction 99.5 ± 1.3% False positive volume fraction 0.1 ± 0.2% Hassanien et al. [38 ] Image classification using PCNN and SVM and using wavelet and fuzzy sets for enhancement MRI Breast cancer detection 70 normal cases, 50 benign and malign cases Hybrid scheme of PCNN and SVM Accuracy SVM: 98% Rough sets: 92% ElNawasany et al. [42 ] Classifying MR images by hybrid perceptron NN MRI Early detection of breast cancer 138 abnormal and 143 normal Perceptron with SIFT Accuracy 86.74%