Systematic Reviews
Copyright ©The Author(s) 2025.
Artif Intell Gastroenterol. Jun 8, 2025; 6(1): 108198
Published online Jun 8, 2025. doi: 10.35712/aig.v6.i1.108198
Table 1 Studies summarizing the role of artificial intelligence in diagnosing gastrointestinal disorders and gastrointestinal malignancies
No.
Title of study
Ref.
Sample size
Validation method
Key limitations
1Artificial intelligence differentiates abdominal Henoch-Schönlein purpura from acute appendicitis in childrenNie et al[15]Paediatric cohortExternalGeneralizability to broader populations unclear
2Application of artificial intelligence in gastroenterologyYang et al[16]Various studiesMixed internal and externalDiversity in study methodologies and validation approaches
3Applications of artificial intelligence in digital pathology for gastric cancerChen et al[17]N/AExternalSmall sample sizes in studies, lack of clinical trial validation
4Applications of artificial intelligence in screening, diagnosis, treatment, and prognosis of colorectal cancerQiu et al[18]N/AMixedLack of long-term validation and diverse population samples
5Artificial intelligence and acute appendicitis: A systematic review of diagnostic and prognostic modelsIssaiy et al[19]8 studiesInternalInconsistent diagnostic performance across studies
6Artificial intelligence technique in detection of early esophageal cancerHuang et al[20]200+ patientsExternalExternal validation required in diverse clinical settings
7Artificial intelligence-assisted system for the assessment of Forrest classification of peptic ulcer bleedingHe et al[21]500+ patientsMixedExternal validation not yet confirmed
8Diagnostic performance of artificial intelligence-centred systems in the diagnosis and postoperative surveillance of upper gastrointestinal malignancies using computed tomography imaging: A systematic reviewChristou et al[22]20 studiesExternalLimited validation on heterogeneous patient populations
9Deep learning for prediction of lymph node metastasis in gastric cancerJin et al[23]1000+ patientsExternalLimited external validation in diverse populations
10Deep learning-based phenotyping reclassifies combined hepatocellular-cholangiocarcinomaCalderaro et al[24]100+ patientsInternalLack of large-scale, multi-center validation
11Deep learning-based virtual cytokeratin staining of gastric carcinomas to measure tumor-stroma ratioHong et al[25]200+ gastric cancer patientsExternalSingle-center validation, lack of long-term clinical data
12Denoised recurrence label-based deep learning for prediction of postoperative recurrence risk and sorafenib response in HCCLi et al[26]150+ patientsExternalNeed for further clinical validation with diverse cohorts
13Development and validation of a real-time artificial intelligence-assisted system for detecting early gastric cancer: A multicenter retrospective diagnostic studyTang et al[27]500+ patientsExternalSingle-region data, limited external validation
14Diagnostic performance of artificial intelligence-centred systems in the diagnosis and postoperative surveillance of upper gastrointestinal malignancies using computed tomography imagingChidambaram et al[28]30 studiesExternalLimited large-scale external validation
Table 2 Studies summarizing the role of artificial intelligence in response prediction and prognostication
No.
Title of study
Ref.
Sample size
Validation method
Key limitations
1Machine learning in gastrointestinal surgerySakamoto et al[31]N/AMixed internal and externalLack of detailed validation cohort; small sample sizes
2A histopathology-based AI system for genetic alteration screening in intrahepatic cholangiocarcinomaXiao et al[32]100+ patientsInternalSmall sample size, lack of external validation
3A nomogram based on a collagen feature SVM for predicting treatment response in rectal cancerJiang et al[33]200+ patientsExternalSingle-center data, limited generalizability
4A novel classification of intrahepatic cholangiocarcinoma phenotypes using MLTsilimigras et al[34]150+ patientsExternalLimited external validation in diverse populations
5AI system for pathologic outcome prediction in early gastric cancerLee et al[36]300+ patientsExternalLack of long-term follow-up data
6AI-based recognition model for colorectal liver metastases in intraoperative ultrasoundTakayama et al[37]100+ patientsExternalSingle-center validation, limited scalability
7Analysis of methionine metabolism and macrophage-related patterns in hepatocellular carcinomaWen et al[38]200+ patientsExternalRequires further clinical validation in diverse cohorts
8AI and ML predicting transarterial chemoembolization outcomesCho et al[39]100+ patientsMixed internal and externalInconsistent prediction performance across centers
9AI based on serum biomarkers predicts efficacy of lenvatinib in hepatocellular carcinomaHsu et al[40]100+ patientsInternalNeeds multi-center validation
10AI-enabled histological prediction of remission or activity in ulcerative colitisIacucci et al[41]150+ patientsExternalLack of long-term clinical outcomes, limited sample size
11AI for lymph node metastasis prediction in gastric cancerYan et al[42]500+ gastric cancer patientsExternalLimited multi-center validation, needs wider cohort testing
12AI in gastrointestinal cancers: Diagnostic, prognostic, and surgical strategiesNagaraju et al[43]N/AN/ALacks detailed validation or large cohorts
13AI in perioperative management of major gastrointestinal surgeriesSolanki et al[44]300+ patientsExternalLimited by single-center studies; needs broader cohort validation
14AI system to determine risk of colorectal cancer metastasisKudo et al[45]200+ patientsExternalValidation needed in multi-center settings
15AI-driven patient selection for preoperative portal vein embolizationKuhn et al[46]150+ patientsExternalRequires larger-scale studies to confirm findings
16Comparison of models for predicting quality of life after hepatocellular carcinoma surgeryChiu et al[47]100+ patientsMixed internal and externalLimited external validation and long-term data
17ML survival framework for pancreatic cancerWang et al[48]300+ pancreatic cancer patientsInternalLack of multi-center validation; small sample size
18Deep learning model for predicting hepatocellular carcinoma recurrenceLiu et al[86]200+ patientsExternalNeed for larger cohort validation
Table 3 Studies summarizing the role of artificial intelligence in predicting recurrence in gastrointestinal malignancies
No.
Title of study
Ref.
Sample size
Validation method
Key limitations
1A machine learning predictive model for recurrence of resected distal cholangiocarcinomaPerez et al[84]654 patientsExternalLimited to a single center, needs multi-center validation
2A novel prediction model for colon cancer recurrence using auto-artificial intelligenceMazaki et al[85]500+ patientsInternalNeeds external validation across diverse cohorts
3Deep learning for prediction of hepatocellular carcinoma recurrence after resection or liver transplantationLiu et al[86]500+ patientsExternalSingle-center validation, needs larger cohort testing
4Deep learning model for predicting gastric cancer recurrence based on computed tomography imagingCao et al[87]200+ patientsExternalNeeds further multi-center validation
5Machine learning model to predict early recurrence of intrahepatic cholangiocarcinomaAlaimo et al[88]100+ patientsInternalRequires external validation for broader applicability
6Prognostic prediction model for elderly gastric cancer patients based on oxidative stress biomarkersZhang et al[89]200+ elderly patientsExternalLack of multi-center data, small sample size
7Consensus machine learning-derived lncRNA signature for stage II/III colorectal cancerLiu et al[90]300+ patientsExternalValidation needed in larger, multi-center trials
8ML identifies autophagy-related genes as markers of recurrence in colorectal cancerWu et al[91]200+ patientsExternalNeeds larger sample size and multi-center validation
9ML models for predicting postoperative peritoneal metastasis after hepatocellular carcinoma ruptureXia et al[92]250+ patientsInternalInadequate external validation across different patient groups
10ML prediction of early recurrence in gastric cancer: Nationwide real-world studyZhang et al[93]1500+ patientsExternalSingle-region data, needs validation in global populations
11CT radiomics and ML predicts recurrence of hepatocellular carcinoma post-resectionJi et al[94]300+ patientsExternalNeeds more validation across different clinical settings
13ML-based gene signature predicts paclitaxel survival benefit in gastric cancerSundar et al[95]350+ patientsExternalNeeds external validation in diverse clinical settings
14CT-based deep learning model for predicting early recurrence in gastric cancerGuo et al[96]200+ patientsExternalNeeds larger multi-center validation
Table 4 Studies summarizing the role of artificial intelligence in liver transplantation
No.
Title of study
Ref.
Sample size
Validation method
Key limitations
1Knowledge domain and frontier trends of artificial intelligence applied in solid organ transplantationGong et al[110]N/AN/ALacks specific cohort data and external validation
2Artificial intelligence and liver transplantation: Looking for the best donor-recipient pairingBriceño et al[111]200+ patientsExternalRequires larger, multi-center validation
3AI for predicting survival following deceased donor liver transplantationYu et al[112]100+ patientsInternalNeeds broader validation in diverse populations
4AI, ML, and deep learning in liver transplantationBhat et al[113]N/AMixedLack of clear sample size and inconsistent external validation
5Bibliometric and LDA analysis of acute rejection in liver transplantationJiang et al[114]N/AN/AThe study lacks clinical validation data and does not include real-world cohorts
6Criteria and prognostic models for hepatocellular carcinoma patients undergoing liver transplantationSha et al[115]150+ patientsExternalNeeds further multi-center validation and more diverse patient groups
7Machine learning in liver transplantation: A tool for unsolved questionsFerrarese et al[116]N/AMixedLack of specific cohort validation, focuses mainly on theoretical models
8ML model for predicting liver transplant outcomes in hepatitis C patientsZabara et al[117]100+ patientsInternalLimited to one patient cohort, needs broader validation
9Machine learning algorithms for predicting liver transplant resultsBriceño et al[118]150+ patientsExternalNeeds validation across different geographical regions and patient types
10Supervised machine learning to predict hepatic immunological toleranceMorita-Nakagawa et al[119]50+ patientsInternalRequires larger cohort size and further multi-center validation
11AI for predicting survival of individual grafts in liver transplantationWingfield et al[120]200+ patientsExternalValidation needed in multi-center studies with diverse populations