Table 4.
Salient Studies of AI and ML in Liver Diseases.
| Study | Disease | Patients | Modality | AI techniques used | Salient features |
|---|---|---|---|---|---|
| Wei et al.16 | HBV/HCV | Train: 343 HBV Test: 147 HBV; 484 HCV |
Clinical and Laboratory data | Decision Tree, Random Forest, Gradient Booster |
Age, AST, ALT and platelet count were used to construct ML algorithms to predict fibrosis in HBV patients. The model was superior to FIB-4 score. |
| Wang K et al.19 | HBV | Training: 266 HBV Validation: 132 HBV |
Ultrasound | Convolutional Neural Network | Deep learning Radiomics of Elastography (DLRE) was superior to 2D-Sheer Wave Elastography and biomarkers for assessing liver fibrosis stages. |
| Konerman MA et al.23 | HCV | Train: 533 HCV Test: 183 HCV |
Clinical and Laboratory data | Logistic Regression, Random Forest |
ML-based longitudinal fibrosis prediction model shows AUROC = 0.78–0.79 for fibrosis progression and AUROC = 0.79–0.86 for clinical progression |
| Vanderbeck et al.27 | NASH/NAFLD | NAFLD-27 Healthy liver-20 |
Pathology data | Support Vector Machine | Automatic classification algorithm of steatosis had an 89% overall accuracy and identified macrosteatosis with ≥95%precision and recall |
| Yip TF et al.30 | NASH/NAFLD | Train: 146 NAFLD;354 Healthy volunteers (HV) Test: 118 NAFLD, 394 HV |
Clinical and Laboratory data | Logistic Regression, Ridge Regression, AdaBoost, Decision Tree |
ML algorithms based on ALT, HDL-C, triglycerides, HbA1C, WBC and Hypertension were used to develop NAFLD prediction scores. Overall accuracy of NAFLD ridge score- 87%, AUROC- 0.87, 92% sensitivity and 90% specificity. |
| Agarwal S et al.40 | CLD/Oesophageal varices | 828 patients having compensate advanced CLD with oesophageal varices (EV) | Laboratory data Endoscopy images Liver stiffness measurement |
Extreme Gradient Boosting (XGBoost) | The accuracy of machine learning (ML)-based model to predict future VB was 98.7 (97.4–99.5)%, 93.7 (88.8–97.2)%, and 85.7 (82.1–90.5)% in derivation (n = 497), internal validation (n = 149), and external validation (n = 182) cohorts, respectively, which was better than endoscopic classification [58.9 (55.5–62.3)%] alone. Patients stratified high risk on both endoscopy and model had 1-year and 3-year bleeding rates of 31–43% and 64–85%, respectively |
| Dong TS et al.43 | Oesophageal varices | Train: 238 Liver cirrhosis Test: 109 Liver cirrhosis |
Clinical and Laboratory data | Random Forest | EVendo score was developed to identify oesophageal varices with AUROC = 0.82, and could spare 30–40% low-risk patients from unnecessary procedures. |
| Minerali et al.74 | DILI | 1036 FDA-approved individual compounds | Biopharmaceutics Drug Disposition Classification System dataset | Bayesian machine learning models | A ML tool named MegaTox™ can predict DILI in early-stage clinical compounds and recently approved FDA drugs. |
| Eaton et al.37 | PSC outcomes | Train: 509 PSC Test: 278 PSC |
Clinical data | Gradient Boosting | A score PREsTo was created using nine variables. This model can predict decompensation and performs better than Mayo Risk score and MELD score. |
| Ahn JC et al.46 | Cirrhosis | Liver cirrhosis: 5212 Age and sex matched controls: 20,728 |
Electrocardiogram (ECG), Clinical data, Laboratory data |
Convolutional Neural Network | AI-Cirrhosis-ECG (ACE) score was created. It has an excellent performance (AUROC = 0.908) for classifying ECGs from patients with cirrhotics and controls. It is positively associated with markers of liver disease specially MELD-Na and trends show improvements after liver transplant. |
| Singal AG et al.56 | HCC risk | Train: 442 Liver cirrhosis Test: 1050 LC |
Clinical and Laboratory data | Decision Tree, Random Forest |
ML algorithm compared to conventional regression models had significantly better accuracy with >80% sensitivity for predicting HCC development |
| Briceño J et al.69 | Post liver transplantation survival | 1003 Liver transplantations (90% train, 10% test) | Clinical and laboratory data | Artificial neural network, Logistic Regression, Decision Tree, Support Vector Machine | 64 donor and recipient variables were used to train ANN to predict LT graft survival and optimize donor-recipient matching. Graft survival prediction was 90.79% with AUROC = 0.82 |
Abbreviations: AI, artificial intelligence; ALT, alanine transaminase; ANN, artificial neural network; AST, aspartate aminotransferase; CLD, chronic liver disease; HCC, hepatocellular carcinoma; HDL, high density lipoprotein; ML, machine learning; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis.