Page 27 - IJAMD-1-1

International Journal of AI
for Material and Design ML in 3D bioprinting of cultivated meat

Table 5. (Continued)
Foods ML method Analytical Sample Details Performance References
technique size
Outputs:
• Adulterant concentration estimate
Beef, Supervised: PLSR FTIR 10 Inputs: R = 0.99 91
2
Meatball • Optimal FTIR spectral region RMSEP = 0.06
in the range of 1250 – 1000 cm
-1
from fat extracted using hexane.
Outputs:
• Adulterant concentration estimate
Abbreviations: BPNN: Backpropagation neural network; DWT-LSTM: Discrete wavelet transform-long short-term memory; FTIR: Fourier
Transform Infrared Spectroscopy; HSI: Hyperspectral imaging; KNN: K-nearest neighbor; LAB: Lactic Acid Bacteria; LBFGS: Limited-memory
Broyden–Fletcher–Goldfarb–Shanno algorithm; LDA: Linear discriminant analysis; LSTM: Long short-term memory; MLP: Multilayer perceptron;
MSI: Multispectral imaging; PC: Principal component; PLS-DA: Partial least squares discriminant analysis; PLSR: Partial least squares regression;
ReLU: Rectified linear unit; RMSE: Root-mean squared error; RMSEP: Root-mean squared error of prediction; SEP: Standard error of prediction;
SIMCA: Soft independent model of class analogies; SVM: Support vector machine; TVC: Total viable count.

total reflectance (ATR) technique. Moreover, the similar Conflict of interest
R value for both calibration and validation datasets in
2
the PLSR models indicates model stability and proves The authors declare that they have no competing interests.
their effectiveness in accurately measuring the extent of Author contributions
pork adulteration within the selected fingerprint region
(1250 – 1000 cm ) (Table 5). Conceptualization: Wei Long Ng
-1
Writing—original draft: All authors
5. Conclusion Writing—review & editing: Wei Long Ng
The advent of cultivated meat stands as a pivotal advancement Ethics approval and consent to participate
in food biotechnology, offering a sustainable alternative
to traditional livestock farming. Cultivated meat mitigates Not applicable.
environmental impacts and aligns with growing concerns for
animal welfare through the direct cultivation of animal cells. Consent for publication
As this technology evolves, it addresses challenges such as cost Not applicable.
scalability and public acceptance, promising a more ethical
and sustainable future for global meat production. Notably, Availability of data
the integration of 3D bioprinting technology can potentially
enhance the precision and complexity of cultivated meat Not applicable.
production. The incorporation of ML approaches into References
cultivated meat production resulted in advanced intelligent
manufacturing that optimizes processes. ML algorithms 1. Post MJ, Levenberg S, Kaplan DL, et al. Scientific,
can predict printing outcomes, characterize meat flavor, sustainability and regulatory challenges of cultivated meat.
and ensure meat quality control, establishing a synergistic Nat Food. 2020;1(7):403-415.
relationship between technology and data. This dynamic doi: 10.1038/s43016-020-0112-z
combination of cultivated meat, 3D bioprinting, and ML holds 2. Tuomisto HL, de Mattos M.J.T. Environmental impacts
immense potential for revolutionizing the meat industry, of cultivated meat production. Environ Sci Technol.
and the prospects for a more efficient, sustainable, and high- 2011;45(14):6117-6123.
quality meat production process become increasingly evident.
doi: 10.1021/es200130u
Acknowledgments 3. Mantihal S, Kobun, R, Lee BB. 3D food printing of as the

None. new way of preparing food: A review. Int J Gastron Food Sci.
2020;22:100260.
Funding doi: 10.1016/j.ijgfs.2020.100260

The study is supported by the NTU Presidential 4. Ng WL, Chua CK, Shen YF. Print me an organ! Why we are
Postdoctoral Fellowship. not there yet, prog. Polym Sci. 2019;97:101145.

Volume 1 Issue 1 (2024) 21 https://doi.org/10.36922/ijamd.2279