Table 3.

Examples of AI applications for process modeling and parameter optimization in 3D bioprinting.

Process category	CMA	CPP	CQA	AI Model	Ref
EBB	Cink	Q, VT, Dnozzle	Shape fidelity	Hierarchical machine learning (HML)	[226]
	$\eta$, $G^{\prime }$	$\Delta P,D_n,V_n\text{,}$$L_n$	Printing resolution	Rheology-informed hierarchical machine learning (RIHML)	[240]
	GelMA composition	Ink reservoir temperature, pressure, speed, platform temperature	Filament morphology, layer stacking	Bayesian optimization	[228]
	–	Air pressure, biomaterial ink temperature, print speed	Print resolution	Bayesian optimization	[227]
	FSA concentration	Nozzle size, printing temperature, pneumatic pressure	Printability	Gaussian process regression (GPR)	[241]
	–	Rb, Rs, Lu, Ll, Rm (nozzle geometrical parameters)	Maximum shear stress	Gaussian process (GP)	[242]
	Material composition	Printing speed, printing pressure, scaffold layer, programmed fiber spacing	Printing quality	RF	[243]
	Biomaterial concentration	Nozzle temperature, printing path height	Printability	SVM	[244]
	Material concentration, solvent usage	Crosslinking mechanism and duration, printer settings, observation duration	Cell viability, filament diameter, extrusion pressure	Support vector regression (SVR), linear regression (LR), random forest regression (RFR), RF, logistic regression classification, SVM	[245]
	Gelatin concentration	Printing speed, flow rate, temperature	Printability, Precision	Fuzzy inference system (FIS)	[246]
	Dilution percentage of bioink	Nozzle pressure, printing speed,	Line width	Fuzzy inference system (FIS)	[247]
	Viscosity, growth factor concentration	Gauge pressure, build orientation, printing speeds	Print resolution	LSTM	[248]
	–	Printing speed, pressure of extrusion, infill percentage	Gel weight, surface area, topographical heterogeneity	SVM, Gaussian model	[249]
	Biomaterial type, biomaterial concentration, crosslinker concentration, cell type, cell number	Crosslinking time, printing pressure, movement speed, nozzle size, cartridge temperature, bed temperature	Cell viability	Bayesian optimization, ANN	[229]
	Material's weight fraction	Extrusion pressure, print speed, z-height	Filament width	Linear regression	[250]
	–	Nozzle temperature, infill density, layer height, printing speed	Tensile strength	Linear regression, RFR, XGB regressor, LGBM regressor, ANN	[251]
	–	Air pressure, biomaterial ink temperature, print speed	The width of printed filament	Bayesian optimization	[227]
	Material's weight fraction	Extrusion pressure, print speed, nozzle diameter, z-height	Filament width	Linear regression	[250]
	–	Layer height, nozzle travel speed, and dispensing pressure	Time, porosity, and geometry precisions	Multi-objective Bayesian Optimization	[252]
	–	Printing speed, extrusion pressure	Width average, width variance height average and height variance	SVM	[253]
	–	Nozzle tip to collector distance	Jet radius profile	GP	[230]
	–	Ratio of the collector speed over the jet speed at the point of interest	Lag distance
	Cell type	Wall shear stress, exposure time	Cell viability	Multi-layer Perceptron (MLP)	[254]
DBB	Viscosity, surface tension	Voltage, diameter of the nozzle	Droplet formation	MLP	[255]
	Viscosity, surface tension	Voltage, nozzle diameter	Droplet deformation	Fully connected neural network (FCNN)	[256]
	Polymer concentration	Voltage, dwell time, rise time	Droplet velocity and volume	Ensemble learning	[257]
	–	Standoff height, applied voltage, ink flow rate	Droplet diameter	Regression analysis (RA), backpropagation neural network (BPNN), neural network trained with genetic algorithm (GA-NN)	[258]
	The type and concentration of solute and solvent	Inner diameter (Din), outer diameter (Dout), the materials of the nozzle and grounded substrate, volumetric flow rate (Q), the distance (L) between needle and grounded substrate, the environmental gas, the applied voltage (V) between the ground electrode and needle	Spraying patterns	ANN,SVM	[259]
	Viscosity (μ), Density (ρ), Conductivity (K), Surface tension (γ), Relative permittivity (κ)	Nozzle internal diameter (Din), nozzle external diameter (Dout), distance between nozzle and grounding electrode (L), applied voltage (V), flow rate (Q)	Droplet diameter	ANN	[260]
	Dimensionless number Z	Rise time, drive'voltage, dwell time, fall time	Drop velocity, drop formation	SVM, KNN, RFs, extreme gradient boosting (XGBoost), MLP	[261]
	Bioink viscosity, cell concentration	Nozzle size, printing time, printing pressure	Droplet size	DT, RF, PageRank, MLP, LSTM	[262]
LBB	–	Digital mask	Printing fidelity	U-Net-like neural network	[224,225]
	–	Digital mask	Printing fidelity	3D U-Net	[263]
	–	Digital mask	Printing fidelity	Convolutional Auto-Encoder (CAE)	[264]
	–	Digital mask	Printing fidelity	Deep neural networks	[265]
	GelMA concentration	UV intensity, UV exposure time, layer thickness	Cell viability	Ensemble learning model	[266]
	–	Exposure time, light intensity, print speed, laser current, laser power, infill density	Young's modulus	ANN	[232]
	Resin viscosity	Cross section size used for synthetic dataset construction, manufacturing velocity, PDMS thickness, constrained surface type, duration of frame, video projection time, groove width, groove depth, cross section size used for separation force boundary construction	Printing success or failure, optimum printing speed	KNN, SVM, decision tree, logistic regression, quadratic discriminant analysis, GP, naiveBayes, ANN, ensemble learning model, Siamese network	[267]