Evaluation of the Quality of Different Brands of Beef Upper Brain
based on Correlation and Principal Component Analysis
Wanyue Hua
1a
, Xiaoning Jiang
2b
, Liting Yan
3c
, Xuefei Hou
1d
and Huijun Liu
1e
1
Food Science and Engineering College, Beijing University of Agriculture, Beijing, 102206, China
2
RDFZ Xishan School, Beijing, 100193, China
3
Beijing Yanxi Yueshengzhai Moslem Food Co.,Ltd., Beijing, 101400, China
Keywords: Beef Upper Brain, Correlation Analysis, Principal Component Analysis.
Abstract: In order to establish the evaluation standard of beef upper brain quality, correlation and principal component
analysis were used to comprehensively evaluate the quality of different brands of beef upper brain.
Conventional methods were used to determine the color difference (Lab) value, colonies number, pH value,
volatile base nitrogen (TVB-N), sensory quality and texture quality indicators of three commercially available
beef upper brain. The results showed that the correlation analysis showed that the colonies number and
chewiness were two key factors that affected the storage quality of cattle upper brain. Principal component
analysis extracted three principal component factors, the first principal component variance contribution was
40.978%, the second principal component variance contribution was 32.993%, the third principal component
variance contribution was 11.893%, and the cumulative variance contribution was 85.863%. The original
complex comprehensive evaluation model of cattle upper brain quality can be replaced by these three principal
components. Using this model to rank the comprehensive meat quality score is: B>A>C. It was expected to
provide technical reference and theoretical basis for follow-up researchers to evaluate the quality of beef upper
brain.
1 INTRODUTION
Beef had many advantages such as high protein, low
fat, rich in minerals, vitamins and a variety of amino
acids. It was an indispensable meat food for ordinary
people on the dinner table (Bai 2020). According to
different processing methods of beef, it can be
divided into three categories (hot fresh beef, cold
fresh beef, frozen beef)0. Cold fresh beef referred to
a low-temperature fresh meat product that rapidly
cools and deacidifies the carcass after slaughtered in
strict compliance with veterinary inspection and
quarantine regulations, and kept it in the range of
0~4℃ during the later processing, transportation and
sales0. From slaughter, processing to marketing, cold
fresh beef had undergone a process of stiffness, de-
rigorization and maturation. During storage, the
protein was normally degraded, Under the strict
a
https://orcid.org/0000-0002-2863-9426
b
https://orcid.org/0000-0003-2165-732X
c
https://orcid.org/0000-0002-5258-7777
control of the food quality and safety management
system, the meat quality, color and elasticity of cold
fresh meat had been improved, and it had become
tender and juicy and has a good taste. At the same
time, the decrease of pH caused lactic acid to inhibit
the reproduction of microorganisms, made cold
Colonies numbers meat safer when eating, and also
prolonged the freshness period0. Cold fresh beef was
first popular in developed countries in Europe and
America, accounting for about 90% of the meat
market circulation. At present, developed countries
such as Europe, America and Japan had more
advanced technology for cold fresh beef processing
technology and circulation technology, and the
quality control and tracking system were also
relatively complete to ensure that the cold beef that
people buy for consumption was safe and reliable0.
At present, there were many indicators for
evaluating beef quality, but there were few studies
d
https://orcid.org/0000-0001-7760-8282
e
https://orcid.org/0000-0001-9527-6728
52
Hua, W., Jiang, X., Yan, L., Hou, X. and Liu, H.
Evaluation of the Quality of Different Brands of Beef Upper Brain based on Correlation and Principal Component Analysis.
DOI: 10.5220/0011164000003444
In Proceedings of the 2nd Conference on Artificial Intelligence and Healthcare (CAIH 2021), pages 52-57
ISBN: 978-989-758-594-4
Copyright
c
2022 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
and evaluations on the quality of beef upper brain, the
weights of various traits were difficult to
scientifically determine, and a comprehensive
evaluation method for beef upper brain quality had
not been established. The principal component
analysis (PCA) was to recombine many indicators
with certain correlations (such as P indicators) into a
new set of independent comprehensive indicators to
replace the original indicators0, this evaluation
method had been widely used in the field of food
quality evaluation. The more the number of indicators
measured by this method and the higher the
correlation between the indicators, the less the
number of corresponding principal components.
Comprehensive evaluation using principal
component analysis had the advantages of
comprehensiveness, comparability, rationality,
feasibility, etc0. Therefore, this article selected three
commercially available beef upper brains to
determine the sensory quality, color difference (Lab)
value, colonies number, pH value, total-volatile basic
nitrogen (TVB-N) and texture quality indicators,
established a principal component analysis evaluation
model for beef upper brain meat quality, extracted
key influencing factors, and through sensory
evaluation to verify the model, in order to provide
technical reference for the evaluation methods and
standards of beef upper brain quality.
2 MATERIALS AND METHODS
2.1 Materials and Instruments
Three kinds of cold fresh beef upper brain samples
were purchased from Hema Xiansheng Supermarket
in Changping District, Beijing, stored in ice packs,
transported back to the laboratory within 30 minutes,
and kept in a refrigerator at 4 ℃ for later use; Plate
Counting Agar Beijing Luqiao Technology Co., Ltd.;
Hydrochloric Acid Beijing Chemical Plant;
Magnesium Oxide, Methyl Red, and
Bromomethylphenol Green Sinopharm Group
Chemical Reagent Co., Ltd.
YXQ-75G Vertical Pressure Steam Sterilizer
Shanghai Boxun Industrial Co., Ltd.; Texture
Analyzer US BROOKFIELD; Testo 205 pH
measuring instrument Testo International Trade
Shanghai Co., Ltd.; CheckMate3 Headspace
Analyzer AMETEK Trading Shanghai Co., Ltd.;
CM-700d Spectrophotometer Shanghai Gaozhi
Precision Instrument Co., Ltd.
2.2 Experimental Method
2.2.1 Sensory Quality Measurement
With reference to the method of Yang Wenting0and
others, this experiment selected four indicators of
color, smell, viscosity, and broth after boiling as the
sensory evaluation indicators of cold fresh beef upper
brain. The total evaluation was divided into the
average value of the sum of the four indicators. The
sensory quality evaluation was performed using a 5-
point system, and the evaluation criteria were shown
in Table 1. Evaluation by 10 professionals.
Table 1: Sensory Evaluation Table.
Evaluation
index
Sensory level
5 points (very good) 4 points (good) 3 points (average) 2 points (bad) 1 point (very bad)
Color
Very bright red and
shiny
Bright red, shiny
Color dark red,
matt
Grayish or pale
color, dull
Dark brown color,
unacceptable
Smell
Has the peculiar smell of
fresh lamb without any
peculiar smell
With the smell of
lamb, no peculiar
smell
Slightly ammonia
smell
Smell of ammonia
Smell of corruption,
unacceptable
viscosity
Moist surface, not sticky
to the touch
The surface is
slightly dry, not
sticky to the touch
Dry surface, moist
cut surface
Dry surface,
slightly sticky
hands
Extremely dry surface,
sticky
Broth after
boiling
The broth is transparent
and clear, and the fat is
agglomerated on the
surface with a fragrance
The broth is more
fragrant and the fat
accumulates on the
surface
The broth has no
fragrance
The broth is
muddy and smelly
Broth is discolored
and has a strong
peculiar smell
Evaluation of the Quality of Different Brands of Beef Upper Brain based on Correlation and Principal Component Analysis
53
2.2.2 Determination of Flesh Color
Expose the sample to 4 ℃ air to develop color for 30
minutes before measurement0, used a calibrated
portable colorimeter to measure the L, a, and b values
of the sample every day, each box of beef upper brain
was randomly measured at 6 sites, and each site was
measured 3 times in parallel, and the average value
was taken.
2.2.3 The Determination of the Colonies
Number Was Determined
According to the method of GB 4789.2 - 2016
"Determination of the colonies number"0.
2.2.4 Determination of pH Value
Used a hand-held pH meter to directly insert the
bovine upper brain sample to determine the pH value
of the product. Each box of samples was randomly
measured at 3 points and the average value was taken.
2.2.5 The Determination of TVB-N Was
Determined
According to GB/T 2009.228 - 2016, the first method
of semi-trace nitrogen determination0.
2.2.6 Determination of Texture Index
The hardness, elasticity and chewiness of fresh beef
upper brain were measured using the texture analyzer
TPA mode. The measurement mode was0: the probe
model was T/46, the moving distance was 5 mm, the
holding time was 2 s, the trigger load was 2 g, the test
speed was 2 mm/s, the recovery time was 5 s, and the
cycle was 2 times. Randomly selected the area and
take the average of 3 measurements.
2.3 Data Processing
Used SPSS 19.0 software (SPSS company) to
perform correlation and principal component analysis
on the experimental data, and the data were all
expressed as the mean±standard deviation of the 3
parallel results.
3 RESULTS AND ANALYSIS
3.1 Correlation Analysis of the Effects
of Different Brands of Bovine
Supramencephalon on Sensory
Quality
Table 2: Pearson correlation analysis of the influence of different brands of cattle on the quality of the upper brain.
L a b pH
TVBN
value
Colonies
number
Hardness Elasticity Chewiness Senses
L 1
a -0.02 1
b 0.278 0.197 1
pH -0.359 -0.5 -0.154 1
TVBN
value
-0.468 -0.182 -0.309 0.583 1
Colonies
numbers
-.687* -0.236 -0.077 0.552 0.826** 1
Hardness -0.187 0.52 0.018 -0.749* -0.096 0.073 1
Elasticity -0.636 0.174 0.015 0.143 0.616 0.823** 0.531 1
Chewiness -0.419 0.445 -0.23 -0.417 0.248 0.383 0.894** 0.766* 1
Senses 0.254 -0.189 -0.204 -0.148 -0.586 -0.69 -0.204 -0.571 -0.336 1
Note: *. Significantly correlated at the 0.05 level (two-sided). **. Significantly correlated at the .01 level (bilateral).
Used SPSS software to analyze the correlation
between the indicators of different brands of cattle on
the brain, The colonies number was significantly
negatively correlated with sensory scores and L
values, and the correlation coefficients were -0.690
and -687 (p < 0.05), which were significantly positive
with the TVB-N value and elasticity, the correlation
coefficients were 0.826, 0.823 (p < 0.01), chewiness
and hardness had a very significant positive
correlation, with a correlation coefficient of 0.894 (p
< 0.01), and a significant positive correlation with
elasticity, with a correlation coefficient of 0.766 (p <
0.05), and a significant positive correlation between
hardness and pH, and its correlation coefficient was -
0.749 (p<0.05) It showed that colonies number and
chewiness were two key factors that affected the
storage quality of cattle upper brain. It can be seen
that the correlation analysis between the indicators
CAIH 2021 - Conference on Artificial Intelligence and Healthcare
54
showed that the information reflected by the
measurement indicators overlaps. Therefore, it was
necessary to perform principal component analysis
on each quality indicator, which helped to improve
the efficiency and accuracy of the comprehensive
evaluation.
3.2 Principal Component Analysis of
Quality Traits of Beef Upper Brain
of Different Brands
The sensory, physical and chemical microbial
indicators of three different brands of beef upper
brain samples were measured, and SPSS software
was used for principal component analysis. The
number of principal components was determined
according to the principle that the cumulative
variance contribution rate reaches more than 85% and
the eigenvalue was greater than 1
0
. From Figure 1,
Figure 2, and Table 3, we can seen that the first
principal component was 3.688, the variance
contribution was 40.978%; the second principal
component was 2.969, the variance contribution was
32.993%; the third principal component was 1.07,
and the variance contribution was 11.893%. The
cumulative contribution rate of the third principal
component had exceeded 85%. Therefore, it was
feasible to use the first three principal components to
evaluate the quality of beef upper brain of different
brands. This showed that the first three principal
components were sufficient to describe each index to
represent the quality of beef brain meat, and the
variance contribution rate of the principal
components was used as the weighting coefficient to
obtain the comprehensive evaluation function
0
. The
comprehensive evaluation function of beef upper
brain quality was obtained: K = 40.978 K
1
+ 32.993
K
2
+ 11.893 K
3
, (K was the number of evaluation
functions, K
1
was the main component 1, K
2
was the
main component 2, and K
3
was the main component
3).
Table 3: Correlation matrix eigenvalues and cumulative contribution rate.
Ingredients
Initial eigenvalue Extract the sum of squares and load
Eigenvalues Variance%
Cumulative
contribution Rate
Total Variance%
Cumulative
contribution Rate
1 3.688 40.978 40.978 3.688 40.978 40.978
2 2.969 32.993 73.971 2.969 32.993 73.971
3 1.07 11.893 85.863 1.07 11.893 85.863
Note: Extraction method-principal component analysis
Figure 1: Component Plot.
Evaluation of the Quality of Different Brands of Beef Upper Brain based on Correlation and Principal Component Analysis
55
Figure 2: Scree Plot.
Table 4: Component loading matrix after principal
component analysis rotation.
Serial
number
variable
Ingredient
1 2 3
X1 L -0.737 0.126 0.287
X2 a 0.093 0.729 0.195
X3 b -0.174 0.178 0.888
X4 pH 0.279 -0.912 0.045
X5
TVBN
value
0.781 -0.442 -0.007
X6
Colonies
number
0.917 -0.342 0.125
X7 Hardness 0.394 0.885 -0.119
X8 Elasticity 0.935 0.175 0.074
X9 Chewiness 0.686 0.659 -0.253
X10 Senses -0.694 -0.007 -0.544
Note: Extraction method-principal component analysis
method. a. Three components have been extracted.
It can be seen from Table 4 that three principal
components were obtained through principal
component analysis, and the expressions of each
principal component were:
Z
1
= - 0.737 X
1
+ 0.093 X
2
- 0.174 X
3
+ 0.297 X
4
+ 0.781 X
5
+ 0.917 X
6
+ 0.394 X
7
+ 0.935 X
8
+
0.686X
9
- 0.694 X
10
；
Z
2
= 0.126 X
1
+ 0.729 X
2
+ 0.178 X
3
- 0.912 X
4
-
0.442 X
5
+ 0.342 X
6
+ 0.885 X
7
+ 0.175 X
8
+ 0.659
X
9
- 0.007 X
10
；
Z3 = 0.287 X
1
+ 0.195 X
2
+ 0.888 X
3
+ 0.045 X
4
- 0.007 X
5
+ 0.125 X
6
- 0.119 X
7
+ 0.074 X
8
- 0.253
X
9
- 0.544 X
10
；
The magnitude of the principal component factor
load represented the contribution rate of the original
variable in the comprehensive variable that was
formed after dimensionality reduction. Therefore, the
factor loading diagram can be used to determine the
main original variables closely related to the principal
components0. According to the absolute value of
each index load, it can be seen that Z
1
mainly
represents elasticity, colonies number, TVB-N value,
L value, Z
2
mainly represents pH hardness and a
value, and Z
3
mainly represents b value.
Table 5: Comprehensive scores and rankings of physical
and chemical indicators of beef upper brain of different
brands.
Brand K Sort
A -0.133 2
B 1.06 1
C -0.933 3
The comprehensive score and ranking of different
brands of beef upper brains were calculated. The
results were shown in Table 5. The higher the score,
the better the quality of beef upper brain. The order
of quality of different brands was as follows: B> A>
C.
CAIH 2021 - Conference on Artificial Intelligence and Healthcare
56
4 CONCLUSIONS
In this study, 10 quality indicators of 3 brands on the
market were measured. Correlation analysis showed
that the colonies number was significantly negatively
correlated with sensory score and L value, and was
extremely significantly positively correlated with
TVB-N value and elasticity; Chewiness had a very
significant positive correlation with hardness, a
significant positive correlation with elasticity, and a
significant positive correlation between hardness and
pH. Further, through principal component analysis,
the dimensionality reduction analysis of 10 indicators
can be used to extract 3 principal components. The
first principal component selected elasticity, colonies
number, TVB-N value and L value; the second
principal component selected pH, hardness and a
value, and the third principal component selected b
value. The cumulative variance contribution rate
reached 85.863%, which can represent large some
indicators, and built an evaluation model: K = 40.978
K
1
+ 32.993 K
2
+ 11.893 K
3
. Using this model to rank
the comprehensive quality of the selected three
different brands of beef upper brains, the results
showed that the best quality variety was brand B.
Based on correlation analysis and principal
component analysis, the comprehensive evaluation of
different brands of beef upper brain quality can
provide theoretical guidance for the evaluation of
beef upper brain quality.
REFERENCES
Bai, J, Wang, H. (2020) Study on the preparation of onion
extract/PVA film and its effect on the quality of chilled
meat. Packaging Engineering, 41(05): 103-108.
Guo, L. Cai, LS. Lin, Z. et al. (2010) Construction of white
tea aroma quality evaluation model based on principal
component analysis. Journal of Tropical Crops, 31(9):
1606-1610.
Liang, L. Du, ARN. Ma, Tao. et al. (2018) Optimization of
pretreatment technology of low-salt cowpea kimchi and
analysis of storage characteristics. Food science,
39(06): 246-251.
Liu, JM. Lin, LJ. Liu, YJ. et al. (2021) Comprehensive
Evaluation of Dry Quality of Different Varieties of
Longan Based on Principal Component Analysis.
Preservation and processing, 21(05) : 127-133.
Liu, Q. (2019) Study on the fresh-keeping effect of cumin,
prickly ashand cinnamon essential oil soncold-fresh
lamb. Gansu Agricultural University.
Mao, Yt. (2017) Research on Governance of Pig Supply
Chain from the Perspective of Quality. Southwest Jiao
tong University.
National Health and Family Planning Commission of the
People's Republic of China. GB4789.2-2010, (2017)
Determination of the total number of bacterial colonies
in food microbiology examination. Beijing: China
Standards Press.
National Health and Family Planning Commission of the
People's Republic of China. GB5009.228-2016, (2017)
Determination of Volatile Basic Nitrogen in Food.
Beijing: China Standard Press.
Wang, H. Chen, L. Chen, K. et al. (2007) Multi-index
comprehensive evaluation method and choice of weight
coefficient. Journal of Guangdong College of
Pharmacy, 23(5): 583-589.
Wang, JJ. Qiao, Q. Ma, LE. et al. (2004) The metabolic
syndrome defined by factor analysis and incident type
2 diabetes in a Chinese population with high
postprandial glucose. Diabe-tes Care, 27:2429-2437.
Wang, P. He, Bb. Li, Zc. et al. (2019) The effect of
biological preservatives on the preservation of chilled
meat. Chinese Journal of Food Science and
Technology, 19(11) : 199-207.
Wu, D. Li, X. (2019) Microbial Contamination and
Preventive Measures in the Process of Chilled Fresh
Meat Preservation. Food Industry Science and
Technology, 40(21): 320-325.
Xu, H. Hao, LZ. Liu, SJ. et al. (2020) Comparison and
analysis of nutritional quality of eliminated female yak
meat in different regions. Food Industry Technology,
41(09): 232-237.
Yang, WT. Li, JL. Kong, F. et al. (2017) Construction of
Frozen Tan Mutton Quality Evaluation Model Based on
Principal Component Analysis. Food Industry
Technology, 38(09): 300-303+313.
Zhao, J. Zhang, T. Liu, Yf. et al. (2018) The effect of long
freezing time on beef quality. Food and Fermentation
Industry, 44(11): 200-209; 215.
Evaluation of the Quality of Different Brands of Beef Upper Brain based on Correlation and Principal Component Analysis
57
