CC BY
6DOI: https://doi.org/10.21323/2414-438X-2022-7-1-16-21
SCIENTIFIC CHALLENGES IN MODELING
/©¡creative ^commons
Available online at https://www.meatjournal.ru/jour Original scientific article Open Access
Received 31.01.2022 Accepted in revised 15.02.2022
MASTICATION OF MEAT USING ENGINEERING TOOLS Accepted or publication 21.02.2022
Ilija V. Djekic
University of Belgrade — Faculty of Agriculture, Belgrade, Serbia Keywords: meat characteristics, meat engineering, FMEA, meat quality Abstract
This paper gives an overview of scientific challenges that may occur while performing modelling meat (as a product) and simulating mastication by using engineering tools. To evaluate these challenges, Failure Mode and Effect Analysis method has been employed to assess six engineering tools often used in analyzing different perspectives of food oral processing. As a result, a risk priority number comprising of severity of the failure, occurrence probability of a failure and difficulty to detect the failure has been calculated. Results show that finite element method and emotion detection are two tools with highest levels of risks. The first method is a known engineering solution used for analyzing different types of materials, but when it comes to meat as a very complex and anisotropic material, risk of inadequate calculations is high. Emotion detection is not so much dependent on meat as a product consumed but on imperfections of software and risk of recognizing false emotions is high. Findings indicate that more research is needed for a more sophisticated use of these engineering tools. Further studies should include other engineering models that simulate meat breakdown during mastication, the role of saliva and jaw movement with the aim to carry out full modelling of mastication of an average meat consumer.
For citation: Djekic, V.I. (2022). Scientific challenges in modeling mastication of meat using engineering tools. Theory and practice of meat processing, 7(1), 16-21. https://doi.org/10.21323/2414-438X-2022-7-l-16-21
Funding:
The results within this research come from a Proof of Concept project No. 5229 "Design of artificial masticator for modelling food oral processing" financed by the Innovation Fund from the budget of the Government of the Republic of Serbia, Ministry of Education, Science and Technical Development through the "Competitiveness and Jobs Project".
Introduction
Modelling of food is very complex due to three constraints: knowledge about a food product, reliability of experimental data and uncertainties associated with food properties [1]. Predictive models and simulations enable development of new scientific approaches and optimizations of food products and food processes [2]. Also, modelling of food behavior can provide information related to food characteristics [3]. This is even more pronounced for multiscale modelling of food tissues with different mechanical properties [4]. Therefore, the main purpose for food engineering is to understand a certain engineering phenomenon combining existing theoretical understanding and available measurements [5].
Meat is considered as a postmortem skeletal muscle tissue of different animals used for human consumption [6]. After slaughter, it undergoes numerous changes, both physiological and biochemical [7]. As a material, it is considered as a matrix comprised of three main elements: muscle fibers, intramuscular connective tissue, and intramuscular fat [8]. However, its standardization is difficult as all these elements depend on a variety of factors such as the species/breed of the animal, age of the animal when slaughtered, type of feeding and other different animal husbandry aspects and finally position of the specific sample in the carcass [9].
Consumption of meat and meat products on a global scale demonstrates two tendencies: (i) an overall rise in consumption mainly caused by the growth of the global population, and (ii) an increase in consumption of meat expressed per capita [10]. Several authors have determined reasons for such a trend, just to mention two most important: dietary habits and nutritional needs for food with animal origin (meat) proteins [11], and sensory enjoyment when consuming meat and meat products [12]. Besides these two trends associated with nutrition and hedonisms, mastication also plays its role in overall perception of meat associated with textural perception) [13,14]. To better understand this quality attribute, it is of utmost importance to understand meat as a material and its mechanical characteristics.
Depending on mechanical properties under load, material science recognizes three types of materials, commonly associated with food: (i) isotropic materials directionally independent; (ii) orthotropic materials (interchangeable across the three main orthogonal axes), and (iii) aniso-tropic with different mechanical properties in all directions [15]. Meat is a very complex system dependent on the interaction between processes and forces of the meat matrix [16]. As such, meat is an anisotropic material, but for different studies/simulations authors consider it as an
Copyright © 2022, Djekic. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/ licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.
orthotropic material [17]. Figure 1 depicts some key words associated with modelling meat.
Figure 1. Word cloud figure associated with meat modelling
Mastication starts from the first bite and ends with swallowing. Food undergoes several phases, from the first bite when incisors initiate the food breakage and first deformation, followed by fractures, bite acceptance in the oral cavity, initial comminution, transportation and distribution of particles inside oral cavity, further comminution, formation of a swallowable bolus and finally swallowing [18]. The latest research confirms that the mastication behavior is more a rhythmic action that creates a pattern and it is dependent on mechanical properties of food rather than its predominant taste [19].
The main objective of this study was to evaluate challenges when modelling mastication of meat using available
Table 1. Severity, Occurrence and Detection rating scale Severity
engineering tools through the Failure Mode and Effect Analysis and as a result weigh potential risks associated with selected tools. As the mastication process has many activities that can be modelled, only the following steps have been investigated: modeling first bite by using results from the following engineering tools — the Warner-Bratzer (BW) test, compression test and Finite Element Method (FEM); modelling bolus characteristics necessary for swallowing using the computer vision system (CVS) for particle size distribution and analyzing the mastication process itself, through video capturing and emotions detection.
Objects and methods
Ranking of risks associated with using engineering tool in modeling meat was performed by professionals with expertise in meat science and food quality, also holding engineering and technological skills. To calculate these risks, the Failure Mode and Effects Analysis (FMEA) has been used as an analytic tool [20]. This technique is very useful as it identifies possible failure modes as well as their causes but in parallel investigates effects of the failures [21]. When using FMEA, it is necessary to use previous knowledge related to similar items or problems [22]. Therefore, it is common to develop an inventory of possible failure modes and evaluate associated risks [23]. For the purpose of this study, a list of potential nonconformities has been populated. The FMEA risk also known as the "risk priority number — RPN" was calculated as follows [21]:
RPN = S x O x D (1)
where:
(S) is the severity of the failure;
(O) stands for occurrence probability of a failure;
(D) stands for difficulty to detect the failure.
Rank Consequence Description
1 None No failure(s)
2 Minor Failure(s) associated with results for one characteristic, not critical-to-quality
3 Low Failure(s) associated with results within one critical-to-quality meat characteristic
4 Major Failure(s) associated with results within more than one critical-to-quality meat characteristics
5 Severe Failure(s) associated with results affecting entire quality of meat
Occurrence
Rank Probability Description
1 Very unlikely Minimal probability of occurrence of failure(s) because offorce majeure
2 Unlikely Occurrence of failure(s) only because of misuse of software / instrument
3 Possible Occurrence of failure(s) only because of errors in previous calculations/estimations
4 High probability Occurrence of failure(s) because of human errors / mistakes
5 Certain Occurrence of failure(s) because of lack of knowledge
Detection Rank Criteria
Very high
High
Low
Remote
Never
Description
Failure(s) associated with results is easily detected
Failure(s) associated with results is detected during initial calculations/estimations Failure(s) associated with results is detected during simulation / validation Failure(s) associated with results is detected during verification No possibility of identifying failure(s) associated with results of modelling
Table 1 outlines weighting factors of the three factors, adopted and modified from [22,24,25]. Experts who participated in the session confirmed that all important nonconformities that might occur while modelling meat were identified. Consensus for each weighting factor was reached with no opposed and/or conflicting opinions linked with final RPN score.
Results and discussion
FMEA Analysis
Results of the FMEA analysis are depicted in Table 2 for the following six selected engineering tools: the Warner-Bratzer (BW) test, compression test, Finite Element Method (FEM), computer vision system (CVS), video capturing and emotion detection.
Warner-Bratzer test
The WB test has been used for many years in order to assess meat tenderness. It measures maximum force calculated as a function of knife movement and compression shear off showing the hardness of meat [26]. As pointed in Table 2, two main issues may occur associated with this test — choice of device parameters and preparation of meat.
When it comes to the choice of test parameters, it has been confirmed that the angle of the cutting edges of the blade may affect results (an increase in shear force), as well as different blade thicknesses and the width between the blade that may influence rupture force values [27]. The RPN value of this issue is 18 mainly as this test is standardized, with the standardized 'Warner-Bratzler' blade used for different texture analysis instruments.
Preparation of a meat sample for testing is of utmost importance. The first criterion is the diameter of the sam-
ple as it must be uniformly round. This is easier to obtain from large animals' muscles with a proposed diameter of 1.27 or 2.54 cm [26]. A culinary method used for preparation is the second criterion [28] followed by chilling to 2-5 °C for obtaining consistency of the material [26]. The final criterion is the direction of fibers as this test has to be performed normally on the axis of the muscle fibers [28]. For this test, the RPN value is 36, mainly dependent on human errors associated with preparation of meat samples.
Compression test
Compression tests are used to assist true stress and strain calculations as proposed by Vallespir et al. [29] and Nieto et al. [30]. In parallel, it enables rupture stress (oR, MPa) and strain (sR) to be extracted from the first peak of the stress-strain curve. Like the WB test, this test also has two main issues that may occur related to the instrument and meat sample.
Opposed to the WB test that may be considered as standardized in terms of the blade characteristics and other instrument parameters, for the compression test a larger number of variables occur such as the test speed, compression percentage, load cell and probe selection. Also, since different mechanical properties associated with texture as a quality characteristic are obtained and/or calculated, the RPN values is 36.
When it comes to meat sample preparation, first, it has to be performed by thin-bladed sharp knives to minimize the damage of the fibers [31]. Second, besides preparation of samples, direction of fibers is in direct correlation with the results [26]. When it comes to 3D modelling, this issue is even more pronounced when it comes to under-
Table 2. Failure Mode and Effect Analysis of modelling meat using engineering tools
No Tool Non-conformity Potential Failure Effect Severity (S) Occurrence (O) Detection (D) Risk
Warner-Bratzler Warner-Bratzler Compression test
Compression test
Finite element method
Inadequate instrument parameters
Wrong direction of fibers
Inadequate instrument parameters
Wrong direction of fibers Inadequate assumptions
Variations in maximum
force value Inadequate reading of maximum force values Variations in values of tested
parameters Inadequate reading of tested parameters Wrong values and inadequate modelling
18 36 36 48 80
1
2
3
Computer vision system
Inadequate color detection Wrong reading of color parameters 3 2 4 24
Inadequate particle preparation Incorrect calculation of particle size distribution 2 4 4 32
4
Video capturing Video clips of low quality Difficulty in oral processing characterization 4 4 3 48
Video capturing Inadequate categorization chews / consumption time Detection of wrong oral processing characteristics 4 2 4 32
Emotion detection Video clips of low quality Difficulty in detecting emotions 5 4 4 80
Emotion detection Inadequate categorization of emotions Detection of incorrect emotions 5 2 3 30
5
standing the direction of compression/expansion related to the fibers [17].
Finite element method
One of the most popular engineering tools is FEM and as such it has found its application in food science and food engineering. This tool enables performing different types of analyses and modelling focused on solving complex mechanical problems [32]. In meat science, its common use is mass/heat transfer [33-35], but with limited application in other dimensions of meat science such as simulating the first bite [17]. When modelling meat is performed using FEM, it is typical to define the shape of the piece (usually as cubic pieces leading to 3D simulation) to enable the use of different software. This is the first assumption when modelling meat. In parallel, other assumptions are usually a type of material (orthotropic for 3D simulation), direction of fibers and direction of forces within the material [17].
To simulate the first bite, the following assumptions are needed: the shape and size of the sample, direction of the first bite related to the direction of fibers, and position of the incisors when initiating the first bite [17]. Also, the following rules apply: (i) WB test values divided by two correspond to the first bite force; (ii) compression test values allow assumption of expansion of meat when subject to specific loading direction and values for this test enable calculation of the Poisson's ratio. These inputs are minimal requirements for mesh construction in FEM using four-node tetrahedral elements [36]. The RPN value for this tool is very high as it is directly dependent on all assumptions and pre-calculations serving as inputs in FEM simulations. Any mistake in pre-calculations and assumptions directly causes wrong values and inadequate modelling.
Computer vision system
CVS is considered as a novel tool used for instrumental evaluation of the meat color [37]. It has advantages compared to traditional colorimeters as latest studies confirm significant differences between L*, a*, b* color values of different types of meat and meat products measured with
CVS opposed to colorimeters traditionally used [16,38]. This equipment also has the potential of being used for particle size distribution analysis, as these high-quality photos enable further computational processing of the number of particles and 2D calculation of the surface area of each particle [28,39].
The use of this tool for color evaluation has a low RPN as this method has been developed and validated [16] and potential failures may be associated with misuse of software for color processing and/or CVS itself. However, for particle size distribution analysis, computational processing of the number of particles and their surface is more dependent on humans (in terms of spreading out the boluses with care in order not to damage the size of the particles, [28]) and consequently RPN has been calculated as 32.
Video capturing
In order to perform oral processing studies for the purpose of calculating the number of chews and consumption time, and calculating different attributes such as chewing cycle duration (s/chew), chewing rate (chews/s), eating rate (g/s) and average bite size (g) [19,40,41], it is common to video capture the mastication process involving human subjects. It is important to position the camera so that the complete upper part of the subject's body is visible and recorded [42]. When video clips are replayed, two potential solutions occur. The first one is the use of a software video analysis that has the feature to analyze graphs of time (x-axis) vs. vertical jaw displacement (y-axis) where chews are visible as peaks [40]. The second solution is the use of humans to count chews while replaying clips and cross checking for accuracy [43]. The latest research on food oral processing confirms that mastication characteristics and behavior from a consumer point of view affects consumer satisfaction in parallel with sensorial properties [19].
Video clips of low quality may cause wrong calculation of key parameters (time/number of chews) and consequently all other oral processing attributes. Therefore, the calculated RPN value is 48. However, if validated software is used, the similar issue may arise, but the RPN is 32 as all potential mistakes and errors are minimized.
Figure 2. Stress distribution during impact of the upper and lower jaw shown for two positions: a) at the edge of the sample; b) at the exact middle of the sample. Colors indicate gradient areas of the stress in the direction of pressure
Emotion detection
It is common to use collected video clips (from oral processing/mastication studies) for analyzing emotions [43]. The first criterion for performing this type of studies is good illumination of the face of the panelists to ensure reliable results [44]. The second one is the categorization of different emotions based on internally developed models using some databases such as DeepFace — face recognition and facial attribute analysis framework [45]. Some publications have detected five types of emotions during mastication — 'neutral, 'angry, 'sad', 'happy', and 'surprise, [43] while others have up to seven — 'angry', 'disgusted', 'happy', 'neutral', 'sad', 'scared', 'surprised' [46]. Finally, before starting these types of studies, it is necessary to avoid all types of biases such as unintended detection of wild facial expression [47] and taking off glasses (if any) as some may mask emotions [46]. Therefore, clear protocol for this type of studies is to have panelists being instructed to look directly into the camera from the first bite to swallowing [43].
Two issues may occur when detecting emotions using video capturing. The first one is in case of low quality of video clips. This issue seriously affects emotion detection and causes incorrect results and as such, the RPN value is 80. The second issue is in case of inadequate categorization of emotions mainly caused by using inadequate software and/or incorrect programing of software. Hence, the calculated RPN value for this failure is 30.
Conclusion
The FMEA-based approach for evaluating risks in using engineering tools needed for modelling meat can provide guidance to meat scientists and food engineers to concentrate efforts on the hot spots that are most i nfluential. Our results recognize the finite element method and emotion detection as two tools with the highest level of risks and tools that are still evolving their industrial and scientific application in meat modelling. Results for the first tool are mainly linked with the complexity of meat as a material and difficulties in modelling, in spite of developed software. On the other side, emotion detection is a promising tool but dependent on the human factor and settings for video capturing of emotions associated with meat consumption and hedonism.
Limitation of this paper is the fact that only six engineering tools have been analyzed associated with the first bite, swallowing of bolus and mastication. Further studies should deploy these (and other) tools for modelling food breakdown from the first bite to swallowing, saliva incorporation, understanding jaw movement and finally modelling the chewing trajectory of an average meat consumer. As the final goal, modelling should utilize all meat changes and bolus breakdown and clearly enable validated simulations in relation with the mechanical and physical properties of all types of meat.
REFERENCES
1. Trystram, G. (2012). Modelling of food and food processes. Journal of Food Engineering, 110(2), 269-277. https://doi. org/10.1016/j.jfoodeng.2011.05.001
2. Carvalho, O., Charalambides, M.N., Djekic, I., Athanassiou, C., Bakalis, S., Benedito, J. et al. (2021). Modelling processes and products in the cereal chain. Foods, 10(1), Article 82. https://doi. org/10.3390/foods10010082
3. Fito, P., LeMaguer, M., Betoret, N., Fito, P.J. (2007). Advanced food process engineering to model real foods and processes: The "SAFES" methodology. Journal of Food Engineering, 83(2), 173185. https://doi.org/10.1016/j.jfoodeng.2007.02.017
4. Abera, M. K., Aregawi, W.A., Ho, Q.T., Rogge, S., Delele, M.A., Am-baw, A. et al. (2016). Multiscale Modeling of Food Processes. Chapter in a book: Reference Module in Food Science. Elsevier, 2016.
5. Djekic, I., Mujcinovic, A., Nikolic, A., Jambrak, A.R., Papade-mas, P., Feyissa, A.H. et al. (2019). Cross-European initial survey on the use of mathematical models in food industry. Journal of Food Engineering, 261, 109-116. https://doi.org/10.1016/j. jfoodeng.2019.06.007
6. Matarneh, S.K., England, E.M., Scheffler, T.L., Gerrard, D.E. (2017). The conversion of muscle to meat. Lawrie' s Meat Science. Elsevier. p. 159-185. . https://doi.org/10.1016/B978-0-08-100694-8.00005-4
7. Bekhit, A.E.-D.A., Carne, A., Ha, M., Franks, P. (2014). Physical interventions to manipulate texture and tenderness of fresh meat: A Review. International Journal of Food Properties, 17(2), 433-453. https://doi.org/10.1080/10942912.2011.642442
8. Spyrou, L.A. (2020). A computational multiscale modeling framework for investigating the mechanical properties of meat. Food Structure, 26, Article 100161. https://doi.org/10.1016/j. foostr.2020.100161
9. Purslow, P.P. (2005). Intramuscular connective tissue and its role in meat quality. Meat science, 70(3), 435-447. https://doi. org/10.1016/j.meatsci.2004.06.028
10. Henchion, M., McCarthy, M., Resconi, V.C., Troy, D. (2014). Meat consumption: Trends and quality matters. Meat science, 98(3), 561-568. http://doi.org/10.1016/j.meatsci.2014.06.007
11. Mathijs, E. (2015). Exploring future patterns of meat consumption. Meat science, 109, 112-116. http://doi.org/10.1016/j. meatsci.2015.05.007
12. Font-i-Furnols, M., Guerrero, L. (2014). Consumer preference, behavior and perception about meat and meat products: An overview. Meat science, 98(3), 361-371. http://doi.org/10.1016/j. meatsci.2014.06.025
13. Ángel-Rendón, S.V., Filomena-Ambrosio, A., Hernández-Car-rión, M., Llorca, E., Hernando, I., Quiles, A. et al. (2020). Pork meat prepared by different cooking methods. A microstructural, sensorial and physicochemical approach. Meat science, 163, Article 108089. https://doi.org/10.1016/j.meatsci.2020.108089
14. Chen, J. (2009). Food oral processing — A review. Food Hydrocolloids, 23(1), 1-25. https://doi.org/10.1016/j.food-hyd.2007.11.013
15. Berthaume, M.A. (2016). Food mechanical properties and dietary ecology. American Journal of Physical Anthropology, 159, 79-104. https://doi.org/10.1002/ajpa.22903
16. Tomasevic, I., Tomovic, V., Milovanovic, B., Lorenzo, J., Dordevic, V., Karabasil, N. et al. (2019). Comparison of a computer vision system vs. traditional colorimeter for color evaluation of meat products with various physical properties. Meat science, 148, 5-12. https://doi.org/10.1016/j.meatsci.2018.09.015
17. Djekic, R., Ilic, J., Tomasevic, I., Djekic, I. (26-29 September 2021). Use of engineering tools in modelling first bite-case study with grilled pork meat. IOP Conference Series: Earth and Environmental Science, 854(1), Article 012022. https://doi. org/10.1088/1755-1315/854/1/012022
18. Ilic, J., Tomasevic, I., Dekic, I. (2019). Modelling solid food oral processing using quality function deployment. Food and Feed Research, 46(2), 227-234. https://doi.org/10.5937/ffr1902227i
19. Djekic, I., Ilic, J., Guillé, R.P.F., Tomasevic, I. (2020). Can we understand food oral processing using Kano model? Case study with confectionery products. Journal of Texture Studies, 51(6), 861-869. https://doi.org/10.1111/jtxs.12550
20. Xiao, N., Huang, H.-Z., Li, Y., He, L., Jin, T. (2011). Multiple failure modes analysis and weighted risk priority number evalu-
ation in FMEA. Engineering Failure Analysis, 18(4), 1162-1170. https://doi.Org/10.1016/j.engfailanal.2011.02.004
21. IEC. (2016). IEC60812:2016 Failure modes and effects analysis (FMEA and FMECA). Commission Electrotechnique Internationale: Geneva, Switzerland.
22. Arvanitoyannis, I.S., Savelides, S.C. (2007). Application of failure mode and effect analysis and cause and effect analysis and Pareto diagram in conjunction with HACCP to a chocolate-producing industry: a case study of tentative GMO detection at pilot plant scale. International Journal of Food Science and Technolog, 42(11), 1265-1289. https://doi.org/10.111Vj.1365-2621.2006.01304.x
23. Papadopoulos, Y., Walker, M., Parker, D., Rüde, E., Hamann, R., Uhlig et al. (2011). Engineering failure analysis and design optimisation with HiP-HOPS. Engineering Failure Analysis, 18(2), 590-608. https://doi.org/10.1016/j.engfailanal.2010.09.025
24. Djekic, I., Pojic, M., Tonda, A., Putnik, P., Bursac Kovacevic, D., Rezek-Jambrak, A. et al. (2019). Scientific challenges in performing life-cycle assessment in the food supply chain. Foods, 8(8), Article 301. https://doi.org/10.3390/foods8080301
25. Djekic, I., Tomic, N., Smigic, N., Udovicki, B., Hofland, G., Ra-jkovic, A. (2018). Hygienic design of a unit for supercritical fluid drying — case study. British Food Journal, 120(9), 2155-2165. https://doi.org/10.1108/bfj-01-2018-0052
26. Novakovic, S., Tomasevic, I. (1-4 October 2017). A comparison between Warner-Bratzler shear force measurement and texture profile analysis of meat and meat products: A review. IOP Conference Series: Earth and Environmental Science, 85, Article 012063. https://doi.org/10.1088/1755-1315/85/1/012063
27. Voisey, P.W., Larmond, E. (1974). Examination of factors affecting performance of the Warner-Bratzler meat shear test. Canadian Institute of Food Science and Technology Journal, 7(4), 243-249. https://doi.org/10.1016/s0315-5463(74)73920-7
28. Djekic, I., Ilic, J., Lorenzo, J.M., Tomasevic, I. (2021). How do culinary methods affect quality and oral processing characteristics of pork ham? Journal of Texture Studies, 52(1), 36-44. https://doi.org/10.1111/jtxs.12557
29. Vallespir, F., Rodriguez, 0., Eim, V.S., Rossello, C., Simal, S. (2019). Effects of freezing treatments before convective drying on quality parameters: Vegetables with different microstructures. Journal of Food Engineering, 249, 15-24. https://doi. org/10.1016/j.jfoodeng.2019.01.006
30. Nieto, A.B., Vicente, S., Hodara, K., Castro, M.A., Alzamora, S.M. (2013). Osmotic dehydration of apple: Influence of sugar and water activity on tissue structure, rheological properties and water mobility. Journal of Food Engineering, 119(1), 104-114. https://doi.org/10.1016/j.jfoodeng.2013.04.032
31. Honikel, K.O. (1998). Reference methods for the assessment of physical characteristics of meat. Meat science, 49(4), 447457. https://doi.org/10.1016/S0309-1740(98)00034-5
32. Yousefi, S., Farsi, H., Kheiralipour, K. (2016). Drop test of pear fruit: Experimental measurement and finite element modelling. Biosystems Engineering, 147, 17-25. https://doi.org/10.1016/j. biosystemseng.2016.03.004
33. Amezquita, A., Wang, L., Weller, C.L. (2005). Finite element modeling and experimental validation of cooling rates of large ready-to-eat meat products in small meat-processing facilities. Transactions of the American Society of Agricultural Engineers, 48(1), 287-303.
34. Santos, M., Zaritzky, N., Califano, A. (2008). Modeling heat transfer and inactivation of Escherichia coli O157: H7 in precooked meat products in Argentina using the finite element method. Meat science, 79(3), 595-602. https://doi.org/10.1016/j. meatsci.2007.12.014
35. Cepeda, J., Weller, C., Thippareddi, H., Negahban, M., Subbi-ah, J. (2013). Modeling cooling of ready-to-eat meats by 3D finite element analysis: validation in meat processing facilities. Journal of Food Engineering, 116(2), 450-461. https://doi. org/10.1016/j.jfoodeng.2012.11.024
36. Wang, L., Sun, D.-W. (2002). Modelling three-dimensional transient heat transfer of roasted meat during air blast cooling by the finite element method. Journal of Food Engineering, 51(4), 319-328. https://doi.org/10.1016/s0260-8774(01)00074-7
37. Tomasevic, I., Djekic, I., Font-i-Furnols, M., Terjung, N., Lorenzo, J.M. (2021). Recent advances in meat color research. Current Opinion in Food Science, 41, 81-87. https://doi.org/10.1016/j. cofs.2021.02.012
38. Tomasevic, I., Tomovic, V., Ikonic, P., Lorenzo Rodriguez, J.M., Barba, F.J., Djekic, I. et al. (2019). Evaluation of poultry meat colour using computer vision system and colourimeter: Is there a difference? British Food Journal, 121(5), 1078-1087. https://doi. org/10.1108/BFJ-06-2018-0376
39. Rizo, A., Peña, E., Alarcon-Rojo, A.D., Fiszman, S., Tárrega, A. (2019). Relating texture perception of cooked ham to the bolus evolution in the mouth. Food Research International, 118, 4-12. https://doi.org/10.1016/j.foodres.2018.02.073
40. Ilic, J., Tomasevic, I., Djekic, I. (2021). Ease of mastication index — Quantification of mastication effort using quality function deployment. Journal of Texture Studies, 52(4), 447-460. https://doi.org/10.1111/jtxs.12621
41. Ilic, J., Tomasevic, I., Djekic, I. (2021). Influence of boiling, steaming, and sous-vide on oral processing parameters of cele-riac (Apium graveolens var. rapaceum). International Journal of Gastronomy and Food Science, 23, Article 100308. https://doi. org/10.1016/j.ijgfs.2021.100308
42. Forde, C., Van Kuijk, N., Thaler, T., de Graaf, C., Martin, N. (2013). Oral processing characteristics of solid savoury meal components, and relationship with food composition, sensory attributes and expected satiation. Appetite, 60(1), 208-219. https://doi.org/10.1016/j.appet.2012.09.015
43. Djekic, I., Ilic, J., Chen, J., Djekic, R., Sotowiej, B.G., Vujadinovic, D. et al. (2021). Analysis of pungency sensation effects from an oral processing, sensorial and emotions detection perspective — case study with grilled pork meat. Applied Sciences (Switzerland), 11(21), Article 10459. https://doi.org/10.3390/ app112110459
44. Rocha, C., Lima, R.C., Moura, A.P., Costa, T., Cunha, L.M. (2019). Implicit evaluation of the emotional response to premium organic herbal infusions through a temporal dominance approach: Development of the temporal dominance of facial emotions (TDFE). Food Quality and Preference, 76, 71-80. https://doi. org/10.1016/j.foodqual.2019.04.001
45. Taigman, Y., Yang, M., Ranzato, M.A., Wolf, L. (23-28 June 2014). Deepface: Closing the gap to human-level performance in face verification. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Article 6909616. https://doi.org/10.1109/CVPR.2014.220
46. Danner, L., Sidorkina, L., Joechl, M., Duerrschmid, K. (2014). Make a face! Implicit and explicit measurement of facial expressions elicited by orange juices using face reading technology. Food Quality and Preference, 32, 167-172. https://doi. org/10.1016/j.foodqual.2013.01.004
47. Ahmed, M.U., Woo, K.J., Hyeon, K.Y., Bashar, M.R., Rhee, P.K. (2018). Wild facial expression recognition based on incremental active learning. Cognitive Systems Research, 52, 212-222. https://doi.org/10.1016/j.cogsys.2018.06.017
AUTHOR INFORMATION
Ilija V. Djekic, PhD, full professor, Department for Food Safety and Quality Management, Faculty of Agriculture, University of Belgrade. Nemanjina 6, Belgrade — Zemun, 11080, Republic of Serbia. E-mail: idjekic@agrif.bg.ac.rs ORCID: https://orcid.org/0000-0002-8132-8299
The author bears responsibility for the work and presented data.
The author declares no conflict of interest.
