Advancements in the diagnosis of biliopancreatic diseases: A comparative review and study on future insights

Table 2 Summary of artificial intelligence-based prediction models for computed tomography scan in clinical studies

Clinical data availability	AI agorithm	Equipment	Reference sandard	Outcome masured	AUC	Ref.
With clinical data	Boruta, gradient-boosting classifier	Siemens, GE	Surgical resection	Residual ALN metastasis	0.866
	Lasso regression	Philips	Surgical resection	SLN metastasis	0.95	[93,94]
	CNN-fast and CNN	GE, Philips	Surgical resection	SLN metastasis	0.817
Without or insufficient clinical data	DCNNs	18FDG-PET/CT (Philips, GE)	Surgical resection	ALN metastasis	0.868
	DA-VGG19	GE, Philips	Surgical resection	ALN metastasis	0.9694
	DT, RF, NB, SVM, ANN	Philips	Surgical resection	ALN metastasis	0.86
	XGBoost	18FDG-PET/CT (GE)	Surgical resection	ALN metastasis	0.89

AI: Artificial intelligence; ALN: Axillary lymph node; ANN: Artificial neural network; AUC: Area under the curve CT: Computed tomography; DA: Deformable attention; DCNNs: Deep convolutional neural networks; DT: Decision tree; RF: Random forest; NB: Naïve Bayes; SVM: Support vector machine; 18FDG-PET: Fluorodeoxyglucose positron emission tomography.