Table 2.

Existing techniques on histopathological images dataset.

Ref.	Disease	Dataset Source	Dataset Type	Dataset Description	Tools	Techniques	Accuracy
[32]	Breast cancer	ICIAR 2018	Histopathological images	1568 images, 249 Bioimaging 2015, 400 ICIAR2018	DNN, CNN, RNN, LSTM	Segmentation, feature extraction, classification	90.5% for 4-class classification task
[42]	Breast cancer	Open source	Histopathological images	BACH2018 (400 images), Bioimaging 2015 (249 images), Extended Bioimaging 2015 (1319 images)	CNN, RNN	Classification K-fold	Single Model 97.5%, Ensemble Model 97.5%, CNN 82.1%
[43]	Breast cancer	ImageNet dataset, ICIAR, ISBI, ICPR, MICCAI	Histopathological images	3771 images	RNN CNN SVM, NVIDIA GPUs	Classification	91.3% for the 4-class classification task
[44]	Breast cancer	Anatomy and Cytopathology Lab, Brazil.	Histopathological images	7909 images	DNN, GCN, softmax classifier	Binary classification	99.44% and 99.01%
[45]	Breast cancer	Wisconsin UCI	Histopathological images	92 images	DNN, RNN	Binary classification	DNN gave better results
[46]	Breast cancer	Not given	Histopathological images	400 images	CNN ML, DL, IHC-Net, Naïve Bayes, SVM and RFD	Segmentation, feature extraction, classification	(98.24%), Ensemble classifier 98.41% F-score and 97.66%