Page 15 - IJAMD-1-1

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

A B

C

D

E
F

Figure 4. The influence of layer thickness on print outcomes is depicted through several scenarios: (A) Ideal layer thickness illustrates the preferred layer
thickness, emphasizing a stress region; (B) reduced layer thickness results in interference and smearing; increased layer thickness leads to (C) elongated,
thin lines and (D) breakage; and (E) different validation points were chosen for the optimization of printing parameters; and (F) an optimal combination
of printing parameters was selected for the printing of cubes and pyramid structures (Scale bar = 2 mm). Figure reproduced from Fu et al. .
45
a bitter taste on its own. Such flavor interactions, standard for sensory analysis, making it an expensive,
48
55
involving various compounds, may be rooted in cognitive time-consuming process that is challenging to implement
interpretations and associations formed through past online for immediate feedback. 56
experiences, rather than solely arising from chemical The integration of ML could facilitate an accurate
49
interactions with taste and smell receptors. correlation between instrumental data and sensory analysis
Texture also plays a role in shaping flavor perception, results, allowing for the prediction of flavor perception with
with a general observation that increased viscosity an appropriate model. ML applications in sensory analysis
corresponds to reduced flavor perception. 50,51 The of meat typically fall into two categories: (i) Predicting
impact varies depending on the thickener used. 52,53 perceived flavor based on molecular composition and
A macromolecular thickening agent affects viscosity (ii) predicting perceived flavor using sensor arrays such as
above a critical coil overlap concentration (c*), and its electrochemical sensors. Recent studies have also evaluated
presence has no impact on flavor perception below this the predictive capabilities of ML in determining whether a
threshold. This observation suggests that changes in given molecular structure will impart a sweet or bitter taste.
54
perceived flavor are unlikely due to molecular interactions
between the thickener and taste or aroma compounds. 3.1. Prediction of meat flavor based on molecular
Consequently, predicting meat flavor from a given set composition
of odor or taste compounds using simple linear models The overall procedures for predicting meat flavor from
becomes challenging. Compounding the complexity is the molecular composition were largely similar. It involves
reliance on trained sensory panels as the current industry the identification of the volatile compounds present with

Volume 1 Issue 1 (2024) 9 https://doi.org/10.36922/ijamd.2279