Table 3. Comprehensive literature review - AI in periodontal diagnosis.

COPD: Chronic obstructive pulmonary disease; NHANES: National Health and Nutrition Examination Survey III; CNN: Convolutional neural network; MLP: Multilayer perceptron; RBNN: Radial basis function neural network; GAN: Generative adversarial networks; IoU: Intersection over Union; PCT: Periodontally compromised teeth.

S.NO	AUTHOR AND YEAR	TITLE	TYPE OF NEURAL NETWORK USED	INFERENCE
1.	Vollmer et al., 2022 [10]	Associations between Periodontitis and COPD: An Artificial Intelligence-Based Analysis of NHANES III	CNN, MLP, RBNN	Deep learning and machine learning algorithms can estimate COPD cases using demographic and dental health characteristic factors, according to study results on an extensive population.
2.	Alotaibi et al., 2022 [13]	Artificial intelligence (AI) diagnostic tools: utilizing a convolutional neural network (CNN) to assess periodontal bone level radiographically—a retrospective study	CNN	The deep CNN algorithm (VGG-16) is useful for detecting alveolar bone loss in periapical radiographs and can accurately assess the level of bone loss in teeth. The results imply that machines can work more efficiently based on level classification and characteristics of captured picture analysis. With further optimisation of the periodontal dataset, it is predicted that a computer-aided detection system will emerge as an efficient technique for assisting in the diagnosis and staging of periodontal disease.
3.	Kearney et al., 2022 [31]	A generative adversarial inpainting network to enhance prediction of periodontal clinical attachment level	CNN, GAN	Artificial intelligence was created and employed to forecast clinical attachment level in place of clinical measurements. A generative adversarial inpainting network with partial convolutions was developed, assessed, and validated to predict clinical attachment level. The inpainting technique was found to be superior to non-inpainted techniques and to be within the 1 mm clinician-determined measurement threshold.
4.	Chifor et al., 2022 [14]	Automatic Segmentation of Periodontal Tissue Ultrasound Images with Artificial Intelligence: A Novel Method for Improving Dataset Quality	R-CNN and U-Net convolutional neural network models	Higher IoU following model retraining using the corrected dataset demonstrated the beneficial effects of the suggested quality check and correction technique by measuring the operator's ground truth segmentation in the 3D space.
5.	Piel et al., 2022 [32]	Artificial Intelligence Aiding In The Periodontal Assessment	CNN	CNN can help the clinician read and identify images more quickly, freeing up more appointment time for services like cleaning and patient education.
6.	Xu et al., 2022 [15]	Evaluation of the Effect of Comprehensive Nursing Interventions on Plaque Control in Patients with Periodontal Disease in the Context of Artificial Intelligence	CNN	In order to increase oral health awareness and give an accurate diagnosis of plaque disease, this study built a convolutional neural network-based oral dental disease diagnosis system for oral care interventions. We persistently and permanently insist on comprehensive daily plaque removal in order to improve patients' physical health and quality of life in those with periodontal disease. We accomplish this by urging patients to take appropriate care of their oral health.
7.	Khaleel and Aziz, 2021 [5]	Using Artificial Intelligence Methods For Diagnosis Of Gingivitis Diseases	Principal Component Analysis (PCA) algorithm, Self-Organizing Map (SOM) algorithm and the Fuzzy Self-Organizing Map (FSOM) network algorithm, Bat swarm algorithm	The BAT is the most accurate method in this study because it had a higher diagnosis accuracy for gingivitis disease equal to (97.942%) in the testing state.
8.	Bayrakdar et al., 2020 [11]	Success of artificial intelligence in determining alveolar bone loss from dental panoramic radiography images	CNN	Periodontal bone loss is successfully detected by the CNN system. In the future, oral physicians may use it to make diagnosis and treatment planning easier.
9.	Farhadian et al., 2020 [16]	A decision support system based on support vector machine for diagnosis of periodontal disease	Support vector machine (SVM)	The best performance was demonstrated by the radial kernel function, which had an overall hypervolume under the manifold (HUM) value of 0.912 and an overall correct classification accuracy of 88.7% when used in the design of the SVM classification model. The findings of the current study demonstrate that the created classification model performs reasonably well in predicting periodontitis.
10.	Lee et al., 2018 [4]	Diagnosis and prediction of periodontally compromised teeth using a deep learning-based convolutional neural network algorithm	CNN	The diagnosis and predictability of PCT could be evaluated with the help of the deep CNN algorithm. The diagnosis and predictability of PCT could thus be evaluated using the deep CNN algorithm. As a result, the deep CNN algorithm could be used to assess the diagnosis and predictability of PCT. Therefore, the deep CNN algorithm could be used to assess the diagnosis and predictability of PCT.