Table 4.

RNN Selected Studies.

Authors Name and Year	Methods	Results	Authors Suggestions/Conclusions
Alarifi and AlZubi, (2018) [51]	MSGSRNN	Accuracy = 99.25%, sensitivity = 97.63%, specificity = 98.28%	Outlined methodology analyzes patient characteristics and aids to know the failure and success rate of the process of implant treatment
Kumari et al., (2022) [52]	M–ResneXt–RNN, HSLnSSO algorithm	Accuracy = 93.67, sensitivity = 94.66, specificity = 92.73, precision = 92.44, FPR = 7.27, FNR = 5.34, NPV = 94.88, FDR = 7.56, F1-Score = 93.54, MCC = 87.35	Difficult to distinguish tiny items and produces rather coarse characteristics.
Singh and Sehgal, (2021) [53]	customized CNN-LSTM	Accuracy = 96%	This model gets lower performance using large datasets.