Refs. | Cancer pathology | Methodology | Features | Database | Performance |
---|---|---|---|---|---|
El Nawar et al. (2019) | Lung cancer | AlexNet (MAN) MAN | Support Vector Machines Principal Component Analysis | X-ray images LIDC-IDRI | Accuracy (97.27%), sensitivity (98.09%) specificity (95.63%), precision (97.80%), F1 score (97.95%) AUC (0.995%) |
Husham et al. (2016) | Â | Wavelet Recurrent Neural Network | Not mentioned | Lung images | sensitivity 93.75%, specificity 66.67%, and accuracy 84% for training data and sensitivity 88.24%, specificity 75%, and accuracy 84% for testing data |
Iftikhar et al. (2017a) | Breast cancer | 4-qubit-quantum circuit with six-layered architecture | Deep features | histopathological dataset | 99% |
Yurttakal et al. (2020) | Â | Multi-layer CNN | Not mentioned | MRI | 98.33% accuracy, 100% sensitivity, 96.9% specificity, 96.55% precision, and 0.0167 loss |
Sharif et al. (2019) | Â | Fusion (RCNN and CNN) | Not mentioned | ICPR 2012, ICPR 2014 [MITOS-ATYPIA-14] histopathology images | 87.6% precision, 84.1% recall and 85.8% F1-measure on ICPR 2012 database, and 84.8% precision, 58.3 recall and 69.1% F1-on ICPR 2014 database |
Escorcia-Gutierrez et al. (2022) | Skin cancer | Ensemble learning and deep learning | Thermal parameters from thermograms of lesions | Infrared thermal imaging | 96.65% precision, 94.11% recall, 95.36% f1-score, 91.85% ROC (AUC) |
Khan et al. (2019) | Â | Neural network (NN) [DSL-1] | Not mentioned | Basal Cell Carcinoma (BCC) | 62% sensitivity and 83% specificity |
Andreeva et al. (2021) | Brain tumour | Support vector machine | Discrete cosine transform-based | MRI and SPECT images | 96.8% accuracy, 95% precision, 94% recall, 93% specificity, 91% F1 score |
Diab et al. (2020) | Â | Bat Algorithm with Fuzzy C-Ordered Means (BAFCOM) Enhanced Capsule Networks (ECN) | Region of Interest (RoI) | MRI | 95.81% accuracy, 94.83% precision, 94.34% recall and 94.85% F1-score |
Al-Koussa et al. (2020) | Â | SVM, KNN, DT classifiers fusion | Adaptive histogram equalization | MRI & SPECT | Accuracy 96.8%, precision 95%, recall 94%, specificity 93%, F1 91% |
Saba et al. (2012) | Â | 3D CNN | Deep features extraction | BRATS MRI | 98.32% highest accuracy |