Machine learning based optimization for novel bio-ink development

Machine learning based optimization for novel bio-ink development

Stať ve sborníku v databázi WoS či Scopus

The quality of bio-ink, the backbone of almost every 3D-bioprinted construct, is one of the most critical aspects of successful 3D bioprinting. Various materials have been successfully used as bio-inks in 3D printing with promising biomedical applications. However, the formulation of printable bio-inks for extrusion-based 3D bioprinting remains a significant challenge in additive manufacturing. Bio-inks must demonstrate good mechanical properties, high biocompatibility, and proper printability to succeed in the chosen application. Moreover, identifying suitable printing conditions for new materials requires time-extensive and resource-demanding experimentation. Most recently, to accelerate bio-ink development, there has been significant attention towards the use of artificial intelligence techniques in this process. Combining machine learning with high-throughput theoretical predictions and high-throughput experiments has altered the traditional trial and error paradigm into a data-driven paradigm. In this study, a novel bio-ink is developed composed of chitosan, gelatin, and agarose using a machine learning method to accelerate the fabrication process. The bio-ink is examined for its printability, rheological properties, hydrophilicity, degradability, and biological response. Rheological analysis displayed that the viscosity of the optimized bio-ink was in a suitable range that facilitated reproducible and reliable printing. Various 3D constructs with different layer orientations were fabricated to test their printability and shape fidelity. The ink was then exposed to mesenchymal bone marrow stem cells (MSCs) to evaluate cell adhesion, growth, and morphology on the surface. The morphological study of the cells showed that they were alive and well grown on the bio-ink. Further characterization using MTT assay demonstrated that cells were still viable on the printed construct after subjecting to physiological conditions for three days. These results suggested that the bio-ink may be a potential biomaterial suitable for use in 3D complex tissue constructs fabrication.

The quality of bio-ink, the backbone of almost every 3D-bioprinted construct, is one of the most critical aspects of successful 3D bioprinting. Various materials have been successfully used as bio-inks in 3D printing with promising biomedical applications. However, the formulation of printable bio-inks for extrusion-based 3D bioprinting remains a significant challenge in additive manufacturing. Bio-inks must demonstrate good mechanical properties, high biocompatibility, and proper printability to succeed in the chosen application. Moreover, identifying suitable printing conditions for new materials requires time-extensive and resource-demanding experimentation. Most recently, to accelerate bio-ink development, there has been significant attention towards the use of artificial intelligence techniques in this process. Combining machine learning with high-throughput theoretical predictions and high-throughput experiments has altered the traditional trial and error paradigm into a data-driven paradigm. In this study, a novel bio-ink is developed composed of chitosan, gelatin, and agarose using a machine learning method to accelerate the fabrication process. The bio-ink is examined for its printability, rheological properties, hydrophilicity, degradability, and biological response. Rheological analysis displayed that the viscosity of the optimized bio-ink was in a suitable range that facilitated reproducible and reliable printing. Various 3D constructs with different layer orientations were fabricated to test their printability and shape fidelity. The ink was then exposed to mesenchymal bone marrow stem cells (MSCs) to evaluate cell adhesion, growth, and morphology on the surface. The morphological study of the cells showed that they were alive and well grown on the bio-ink. Further characterization using MTT assay demonstrated that cells were still viable on the printed construct after subjecting to physiological conditions for three days. These results suggested that the bio-ink may be a potential biomaterial suitable for use in 3D complex tissue constructs fabrication.

Bio-ink development, machine learning-based optimization, extrusion 3D bioprinting

Bio-ink development, machine learning-based optimization, extrusion 3D bioprinting

HASHEMI, A.; EZATI, M.; VIČAR, T.; ZUMBERG, I.; CHMELÍKOVÁ, L.; ČMIEL, V.; PROVAZNÍK, I.

2026

12.06.2022

Singapore

978-3-032-20290-1

Proceedings of IUPESM World Congress on Medical Physics and Biomedical Engineering XXVII

IFMBE Proceedings

140

Francouzská republika

1

13

13