Table 1.

A review of selected studies.

Study	Technique(s)	Results	Limitations
Chicco and Jurman (7)	RF classifier model	The [−1, +1] interval of MCC has increased by 8.25%	The validation cohort dataset lacked some of the attributes of the discovery cohort dataset.
Kashif et al. (8)	K Nearest Neighbor, kStar, Bayesian Network, Randomized Forest, Radial Basis, PART, Logistic Regression, OneR, Svms, and Multi-Layer Perceptron	Acc: 87%	Lack of data balancing techniques
Panigrahi et al. (11)	Web-based Expert System Shell	Knowledge base consists of 59 rules to design the expert system	Procedural knowledge can be enhanced for more effective diagnosis
Wicaksno and Mudiono (12)	certainty factor was used for early diagnosis of hepatitis	CF = 97%	Limited rule base
Wu et al. (13)	DeepHBV model	AUROC = 0.6363 AUPR = 0.5471	Lack of appropriate hidden layer selection
Butt et al. (14)	Intelligent Hepatitis C Stage Diagnosis System	Precision (94%)	Lack of external validation
Orooji and Kermani (15)	machine learning to handle unbalanced data in hepatitis diagnostics	More than 90%	Skewed dataset
Parisi and RaviChandran (16)	Merges neighborhood component analysis and ReliefF Lagrangian SVM Classifier (LSVM)	F1-score = 94%	Expanding its applicability to additional hematological diseases in order to improve patient outcomes more comprehensively.