Page 8 - IJB-7-1

Application of Machine Learning in 3D Bioprinting
and virtual shadow. Creating such cell-level twins of Interestingly, ML had been applied to nanotechnology
organs requires high quality tissue specimens and advanced to develop nanocomputing hardware that can boost
imaging and 3D reconstruction methods. Fortunately, the artificial-intelligence-based applications . It could be
[29]
Human BioMolecular Atlas Program from Institute of a reciprocal advancement to expect in future. Another
Health in the United States , which aims to develop an topic worth watching is ML-based programmable design
[23]
open and global framework to create 3D molecular and for 4D printing , as it is relevant to 4D bioprinting, a
[30]
cellular atlas of the human body, may enable the building method in which bioprinted tissues transform shape, size,
of an integrated tissue map across scales. However, as or pattern over time . Aside from academy, the industry
[31]
pointed out in Campos and De Laporte , digital tissues also expects a bright future for use of artificial intelligence
[8]
should not only enable architectural replication of native in 3D bioprinting. For example, in 2019 Procter and
tissues but also be biologically functional. This would Gamble partnered with a biotechnology company Aether
require the capability of assigning fidelitous tissue to develop AI 3D Bioprinter .
[32]
materials to the digital twin and a profound understanding
of individual and collective behaviors of cells. Cell- 6. Toward digital bioprinting
based mathematical models and software , which have Application of ML in bioprinting and biofabrication
[24]
been extensively used in computational biology, may be will induce dramatic transformation and bioprinting
useful tools for modeling cell and tissue properties and will became a part of digital industry and information
behaviors to enable the simulation of biological functions technology [33,34] . What could be done to implement
of the virtual tissue and organs. In fact, from the economic these forthcoming transformations? First, bioprinting
point of view, an alternative but efficient method should community must attract experts in computer sciences,
be one that directly converts current magnetic resonance mathematical modeling, computer simulations, and ML.
imaging (MRI)/confidence interval-based 3D models into Second, special efforts must be done for generation,
cell-based models, that is, cells and tissue properties are assembly and maintaining of desirable Big Data.
intelligently assigned to a virtual organ model with spatial Maintaining and up-dating of such databases are essential.
accuracy and material diversity by artificial intelligent Third, digital organ twins based on sophisticated
algorithms. Slicing of the digital twin for layer-by-layer mathematical modeling and advanced software
bioprinting should also be intact cell-based and matching will become a new type of knowledge presentation,
extrusion layer thickness, which is very different from accumulation, and compaction in bioprinting. Finally,
common slicing in 3D printing. Another alternative further during transition from empiric to digital approach
empowering our imagination is in vivo cellular imaging bioprinting will enter in digital era and it will become not
such as MRI , which can map the anatomic locations of descriptive but rather predictive technology increasingly
[25]
specific cells within living tissue. Given that ML has been based on virtual or in silico experiments.
successfully used for recognition of cell phenotype , it
[26]
might be reasonable to imagine “in vivo 3D scan” of a 7. Conclusion
patent-specific live tissue model into a digital twin with
cellular resolution. Nevertheless, the immediate impact of In our opinion, when applying ML to bioprinting the
the digital twins of organs on bioprinting is that the in vivo most important and urgent challenges are: (1) To build
performance of physical bioprinting such as preclinical training databases for ML from Big Data curation and
as well as clinical studies must collect information with (2) to build digital twins of human tissue/organs. The
specially designed assay, biosensor, and so on for updating goal is to achieve a predictive power of digital twin of
original model in the form of digital twin. Furthermore, human tissue/organ based on Big Data which is close to
the cell-level digital twins together with physical 3D virtual crash test in automobile industry. Ultimately, we
bioprinting could also revolutionize biology fundamentally hope to see a standard bioprinting simulation practice in
by building tissues from scratch to explore entirely future to reduce or replace present 3D bioprinting studies.
new cell configurations for cell cross-talks and cellular We envision that future 3D bioprinting will become more
morphogenesis . This would help provide new insights digital and in silico, and eventually strike a balance
[27]
into the challenging question: “Print me an organ! Why we between virtual and physical experiments to maximize the
are not there yet?” which was recently raised in Ng et al. . efficiency of bioprinting resource utilization. In future,
[28]
digital bioprinting will become a new growth point for
5. Other aspects of the future digital industry and information technology.
In future, it is necessary to include the development of References
correspondent infrastructures such as education and
training specialists and development and adaptation 1. Yu C, Jiang J, 2020, A Perspective on Using Machine
of software and computational power and so on. Learning in 3D Bioprinting. Int J Bioprinting, 6:95.

4 International Journal of Bioprinting (2021)–Volume 7, Issue 1