Page 23 - manuscript_ijb05596

2.4. Optimization of 3D printing parameters

In addition to the crosslinking method, the optimization of 3D printing parameters
was also crucial to the performance of hydrogels. In recent years, artificial intelligence

(AI) technology has been introduced to efficiently optimize the printing process.

Precision optimization of 3D bioprinting parameters, including temperature, extrusion

pressure, and cross-linking kinetics, is essential for fabricating hydrogel scaffolds with

59
high structural fidelity and uniform material properties . Traditional parameter
optimization in 3D bioprinting depended heavily on operator experience and numerous

time-consuming experiments, resulting in inefficient processes that were challenging

60
to standardize. Sakib Mohammad et al. accurately predicted the rheological properties
of 3D-printed polyacrylamide hydrogels using deep learning models and inferred

multiple feasible material composition and printing parameter combinations from the

target modulus values through generative AI models. These AI methods have

significantly reduced the cost of experimental trial and error, achieving a forward

mapping from material formulation to performance prediction and a reverse design
from performance requirements to formula generation. Furthermore, the advancement

in 3D printing and AI technologies had enhanced material precision and adaptability,

showing that traditional diabetes treatment strategies may soon become insufficient to

61
meet emerging therapeutic demands. Zhou Jianhua et al. developed an AI-assisted
high-throughput system (AI-HTPCSS) combining a pneumatic extrusion 3D bioprinter

with AI image analysis to rapidly optimize 3D printing parameters for tissue

engineering applications. Based on AI-HTPCSS, the printing conditions of the hydrogel

architecture with uniform structures were screened in a high-throughput manner. Hyun-

62
Do Jung et al. developed a functional 3D printing ink composed of salmon sperm
DNA and sponge-inspired DNA-induced biosilica for machine learning-based 3D

printing of wound dressings (Figure 5). These biomimetic 3D printed hydrogels

prepared by DNA-induced biomineralization strategy provided excellent functional

platforms in the repair of acute and chronic diabetic wounds, demonstrating the broad

22