Page 150 - v11i4

International Journal of Bioprinting AI for sustainable bioprinting

Figure 5. Machine learning in optimizing processing parameters for bioprinting. (A) A hierarchical machine learning model used for optimizing three-
dimensional printing parameters. (i) The schematics of the model integrating laboratory-controlled system variables, middle-layer physical relationships,
and statistical inference to predict and enhance print fidelity based on dimensional error metrics. (ii) Visualization of trade-offs in optimizing printing
parameters for high-fidelity features. Reprinted with permission from Bone et al. Copyright © 2020, American Chemical Society. (B) Artificial
85
85
intelligence (AI)-assisted high-throughput printing-condition-screening system. Real-time image recognition analyzes printed patterns, classifying them
into droplets, droplet lines, or continuous lines. AI-generated phase diagrams guide parameter selection for achieving optimal print fidelity and uniformity,
reducing material waste and improving sustainable bioprinting efficiency. Reprinted with permission from Xu et al. Copyright © 2020, Springer Nature.
88
88
(C) Machine learning-guided parameter optimization in bioprinting. (i) Low-printability validation: Parameters resulted in (A-1) irregular filaments, (A-2)
deformed cube, and (A-3) non-uniform grid. (ii) High-printability validation: Optimized parameters produced (B-1) uniform filaments, (B-2) well-formed
cube, and (B-3) geometrically accurate pyramid. Scale bar: 1.87 mm for (A-1) and (B-1); 2 mm for (A-3), (B-2), and (B-3). Reprinted from Fu et al. 87
87
Volume 11 Issue 4 (2025) 142 doi: 10.36922/IJB025170164