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

Table 1 Comparative overview of proral colangioscopy, endoscopic retrograde colangiopancreatography, and edoscopic utrasound

Aspect	Endoscopic utrasound	Endoscopic retrograde cholangiopancreatography	Peroral colangioscopy
Technique involved	Combines endoscopy and ultrasonography	Combines endoscopy and fluoroscopy; use of contrast dye and radiography	Insertion of an endoscope via the mouth using advanced imaging
Purpose	Primarily diagnostic	Diagnostic and therapeutic	Detailed diagnostic imaging and therapeutic interventions
Procedure	Use of an endoscope with an ultrasound probe for internal imaging	Injection of contrast dye into the ducts, with radiographic images taken with real-time guidance	High-resolution visualization of the bile and pancreatic ducts
Imaging quality	High-resolution ultrasound imaging	Real-time fluoroscopic guidance	High-resolution; detailed visualization
Technology	Ultrasound-guided fine-needle aspiration biopsy	Fluoroscopy for real-time imaging	Often incorporates digital and high-resolution imaging systems
Primary clinical uses	Pancreatic cancer detection and staging	Diagnosing and treating bile duct obstructions	High-resolution imaging of the bile and pancreatic ducts
	Chronic pancreatitis and biliary disease evaluation	Gallstone removal, stent placement, and stricture dilation	Identifying small lesions and ductal changes
	Evaluation and sampling of submucosal lesions	Stricture and tumor management	Stone removal, stent placement, and dilation of strictures
Advantages	Minimally invasive with high-resolution imaging	Combined diagnostic and therapeutic capabilities	Enhanced imaging quality
	Guided biopsies, including extraluminal targets	Immediate symptom relief and treatment	Reduced radiation exposure
	Ability to reach and biopsy beyond the GI tract	Proven efficacy with a high success rate	Improved diagnostic accuracy via digital innovations
Risks and limitations	Procedure-related risks (e.g., bleeding, infection, and perforation)	Higher rates of complications (e.g., pancreatitis, infection, and bleeding)	Technically demanding; requiring specialized training
	Complementary to ERCP in therapeutic procedures	Radiation exposure from fluoroscopy	Operator dependency affecting outcomes
	Technically demanding	Technological limitations based on the equipment	Anatomical challenges in accessing the ducts
Patient selection	Excellent for staging, lesion assessment, and biopsies	Ideal for immediate therapeutic intervention during diagnosis	Useful for detailed diagnostic evaluations
	Complementary to ERCP in addressing limitations	Suitable for several biliary and pancreatic conditions	Challenges with a complex anatomy
Therapeutic role	Complementary to ERCP in therapeutic procedures	Notable therapeutic capabilities (stone removal, stenting)	Stone removal, stent placement, and dilation
Biopsy capability	Combines endoscopy with ultrasonography	Can collect small tissue samples (biopsies)	Can be performed under direct visualization
Invasiveness	Primarily diagnostic	More invasive with a higher risk of complications	Less invasive than surgery
Imaging vs therapeutics	Endoscope with an ultrasound probe for internal imaging	Balanced diagnostic and therapeutic functions	Useful for high-resolution imaging of small lesions and ducts
Complications	High-resolution ultrasound imaging	Higher risk of pancreatitis, infection, and perforation	Risk of infection, bleeding, and perforation

ERCP: Endoscopic retrograde cholangiopancreatography.

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.

Table 3 AI-Assisted based prediction models for magnetic resonance imaging models

Clinical data availability	AI algorithm	Equipment	Reference standard	Outcome measured	AUC	Ref.
With clinical data	SVM	1.5 T GE	Surgical resection	ALN metastasis	0.87
	SVM	3.0 T GE	Surgical resection	ALN metastasis	0.810
	RF	N/A	Surgical resection	ALN metastasis	0.91
Without or with insufficient clinical data	LDA, RF, NB, KNN, SVM	3.0 T Siemens	FNA or surgical resection	ALN metastasis	0.82
	SVM, KNN, and LDA	3.0 T Siemens	FNA or surgical resection	ALN metastasis	0.8615
	LDA	1.5 T Aurora	Surgical resection	ALN metastasis	0.812
	SVM, XGBoost	3.0 T GE	Surgical resection	ALN metastasis	0.83
	SVM	1.5 T Philips	Surgical resection	SLN metastasis	0.852
	CNN	1.5 T GE	18FDG-PET	ALN metastasis	0.91
	RF	1.5 T Philips	Surgical resection	SLN metastasis	0.868
	Lasso regression	1.5 T Siemens	Surgical resection	ALN metastatic burden	0.81

AI: Artificial intelligence; ALN: Axillary lymph node; ANN: Artificial neural network; AUC: Area under the curve; CNN: Convolutional neural network; DA: Deformable attention; DCNNs: Deep convolutional neural networks; DT: Decision tree; FNA: Fine-needle aspiration; KNN: k-nearest neighbors; LDA: Linear discriminant analysis; NB: Naïve Bayes; RF: Random forest; SLN: Sentinel lymph node; SVM: Support vector machine; 18FDG-PET: Fluorodeoxyglucose positron emission tomography.

Table 4 Summary of key biomarkers and their diagnostic performance

Biomarker	Primary use	Sensitivity	Specificity	Detection method	Clinical applications	Limitations
CA 19-9	Pancreatic cancer	80%-90%	70%-80%	Enzyme-linked immunosorbent assay (ELISA)	Used in monitoring disease progression and treatment response	Elevated in benign conditions; lacks specificity
KRAS mutations	Pancreatic cancer	High	High	Polymerase chain reaction (PCR); next-generation sequencing (NGS)	Identifies high-risk patients, guides targeted therapies	Limited sensitivity in early-stage cancer
Amylase/lipase	Acute pancreatitis	> 90%	70%-80%	Serum biochemical assays	First-line test for diagnosing acute pancreatitis	Cannot distinguish between acute and chronic cases
Alpha-fetoprotein	Hepatocellular and biliary carcinoma	60%-70%	80%-90%	ELISA, chemiluminescent immunoassay	Used in screening for hepatocellular carcinoma	Limited specificity in biliary malignancies
MicroRNAs (miR-21, miR-196a)	Early detection of pancreatic cancer	85%	90%	Reverse transcription PCR (RT-PCR); RNA sequencing	Potentially noninvasive biomarker for early detection	Requires further validation and standardization

CA 19-9: Cancer antigen 19-9.