Table 6.

The existing DL models on implant and their functions and performance.

Authors	DL Models	Year	Training Dataset	Validation/Test Dataset	Functions	Best Performance of DL	Time- Consuming
Khajeh et al. [70]	CNN	2019	620	54 for validation 43 for test	Bone density classification	Accuracy: 0.991 Precision: 0.952	76.8 ms
Lin et al. [71]	Nested-U-Net	2022	605	68	Bone density classification	Accuracy: 0.91 DSC: 0.75	No
Yi et al. [72]	QCBCT-NET	2021	200		Bone mineral density measurement	Pearson correlation coefficients: 0.92	No
Saeed et al. [73]	CNN	2022	350	100 for validation 50 for test	Missing tooth regions detection	Accuracy: 0.933 Recall: 0.91 Precision: 0.96 F1 score: 0.97	No
Shumilov et al. [74]	3D U-Net	2021	75		Bone height\thickness\canals, missing tooth, sinus measuring	Right detection Canal: 0.722 Sinuses/fossae: 0.664 Missing tooth: 0.953	No
Chen et al. [75]	CNN	2022	2920	824 for validation 400 for test	Perioperative plan	ICCs: 0.895	0.001 s for DL 64~107 s for manual work
Wang et al. [76]	CNN	2022	1000	150	Implant stability	Precision: 0.9733 Accuracy: 0.9976 IoU: 0.944 Recall: 0.9687	No