ABSTRACT
This study demonstrates the antioxidative stress potential of
Shanyao–Fuling–Yiyiren (SFY) decoction—a Chinese polyherbal formulation
derived from Si Fang decoction—by establishing a systematic framework
that integrates network pharmacology, molecular docking, in vitro
synergy assays, cellular experiments, and 3D printing. Despite its long
traditional use, the molecular and cellular mechanisms underlying its
antioxidative effects remain unclear, and its formulations are based
more on empirical methods than on systematic design. To fill this gap,
a fuzzy mathematical model was used to select the optimal polyherbal
combination. A central composite circumscribed design determined that a
Shanyao:Fuling:Yiyiren ratio of 2:2:1 maximized radical scavenging,
with a strong correlation (R ^2 = 0.9665) between antioxidant activity
and the combination index. Furthermore, network pharmacology, molecular
docking, and cell‐based assays jointly confirm the AKT1/GSK3β/HIF1α
pathway plays a crucial role in preventing the antioxidant effects of
SFY. Finally, the development of 3D printing SFY‐inks with the
optimized shape fidelity suggests promising applications for both
nutraceuticals and hepatocellular carcinoma diagnosis. Overall, the
results prove that 3D printing SFY‐based polyherbal formulation with
promising antioxidant potential and maximum synergism may indeed be a
potential source of preventing oxidant damages in pharmaceutical and
food industries.
Keywords: 3D printing food, AKT1/GSK3β/HIF1α pathway, fuzzy
mathematical model, network pharmacology, Shanyao–Fuling–Yiyiren
(SFY)‐based polyherbal formulation
__________________________________________________________________
SFY‐based polyherbal formulation was selected based on fuzzy
mathematical model. The best combinations with synergistic effect were
optimized and identified by central composite circumscribed design.
Network pharmacology and molecular docking‐based approach revealed that
the potential mechanisms of decoction against oxidant damage via
multicomponent, multitarget, and multipathway. Decoction presented
ideal antioxidative protection activity against H[2]O[2] damage.
Decoction‐based inks have great practical printability and 3D printing
performance for potential application in functional food production.
graphic file with name FSN3-13-e70349-g005.jpg
1. Introduction
About 95% of humans worldwide stay up late at night, but handling jet
lag can disrupt circadian rhythms, which is considered a stress state
related to many diseases, including metabolic disorders and liver
cancer. As a vital organ in charge of detoxification, the liver is very
sensitive to damages from physiological stress, resulting in
life‐threatening conditions (e.g., liver injury) with a high mortality
rate and poor prognosis. Among the many causes of liver injury,
uninterrupted production of reactive oxygen species (ROS) from both
endogenous and exogenous origins has received much attention, including
hydroxyl radical (OH˙), hydrogen peroxide (H[2]O[2]), hydroxyl ion
(OH^−), singlet oxygen (^1O[2]), superoxide anion (O[2]˙^−), and ozone
(O[3]) (Chen et al. [44]2020). Studies have demonstrated a connection
between prolonged exposure to ROS and liver diseases (Prieto and
Monsalve [45]2017). The elevated levels of ROS are intracellular
signaling factors that remarkably enhance hepatic stellate cell
activation and extracellular matrix generation amid liver injuries.
Additionally, ROS activates significant oncogenic pathways leading to
hepatocarcinogenesis, such as extracellular signal‐regulated kinase,
protein kinase B, c‐Jun N‐terminal kinase, hypoxia‐inducible factor
(HIF), microtubule‐related protein kinase, and intensified cellular DNA
mutations (Hammouda et al. [46]2020). Concurrently, concerns about the
adverse effects of synthetic antioxidants in foods have diverted more
attention to novel sources of natural antioxidants from the safety
aspect. Given the critical role of oxidative stress in liver disease
occurrence, the role of phospholipid peroxidation in liver injuries
shall be urgently explored. Against this backdrop, we hypothesize that
phospholipid peroxidation represents a key mechanistic link between
circadian disruption and stress‐related liver injuries.
Emerging evidence has described that traditional medicinal herbs
potentially contribute to preventing oxidative stress‐related chronic
diseases (Allison et al. [47]2025; Ashraf et al. [48]^2024). These
herbs have been applied conventionally for millennia by many cultures
as medicine, flavoring reagents, and even food preservatives, and are
basically cheap and available for poor populations. Researchers have
tested their antioxidant abilities and potential replacements of
synthetic additives in protecting food and cosmetic products from
oxidative damages. Specifically, Shanyao (Rhizoma Dioscoreae), Fuling
(Poria cocos (Schw.) Wolf.), and Yiyiren (Coicis Semen) are widely used
in traditional medicine, owing to their potent antioxidant properties.
Shanyao in the crude form has long been used as a spice, dietary
supplement, and a constituent of many traditional Asian medicines (Luo
et al. [49]2024). Shanyao contains curcumin, which has these
pharmacological abilities owing to its basic beneficial antioxidant,
anti‐inflammatory, antibacterial, and anticancer abilities (Alam
et al. [50]2024). Fuling has diverse pharmacological activities against
rheumatoid arthritis, type II diabetes, multiple sclerosis,
atherosclerosis, Alzheimer's disease, and other chronic diseases (Guo
et al. [51]2025). Free radicals, which are key stimuli for
carcinogenesis, can be inhibited by Fuling from modulating lipid
peroxidation of membranes or oxidative DNA harms. As for the
antioxidant role, Yiyiren is proved to effectively scavenge diverse
risky free radicals, including ROS, NO[2] radicals, O[2]˙^−, and OH˙
(Zhang et al. [52]2024).
Regardless of the abundant research, observations about the antioxidant
synergism in mixtures are still deficient. Moreover, recent studies
have raised significant controversy regarding the standardization of
such evaluations, reflecting a wider gap in understanding the
underlying mechanisms of antioxidant synergism (Cnudde et al. [53]2024;
Eawsakul and Bunluepuech [54]^2024; Shen et al. [55]^2025). This lack
of consensus has led to speculation and explanations in the literature,
many of which lack empirical verification. In view of antioxidant
synergism effects, the mixing of antioxidant extracts may induce a
synergism to generate a better antioxidant effect than the sum produced
by single extracts (additives) (Eawsakul and Bunluepuech [56]2024;
Kurnia et al. [57]^2025). Moreover, the ratio and type of individual
herb extracts are pivotal in deciding the chemo‐preventive potential or
antioxidant capacity in a health‐benefiting herb combination. Thus, the
regulation of antioxidant properties shall be investigated according to
the proportions of herbs in a mixture that can be used to develop
functional foods and pharmaceutical products at different contents. For
instance, the combination of Osmanthus fragrans flower extract with
four types of tea (Longjing, Tieguanyin, black, or Pu'er Tea) can
synergistically scavenge 2,2‐diphenyl‐1‐picrylhydrazyl free radicals
(Mao et al. [58]2017), demonstrating how specific combinations can
enhance antioxidant performance beyond individual components.
To our knowledge, though the antioxidant activities of Shanyao, Fuling,
and Yiyiren have been extensively reported, there is little research on
the bio‐effects of extracts combined with bioactive compounds, which
are believed to improve the antioxidant benefit of free radical
scavenging. Therefore, this study was aimed primarily to explore the
antioxidant interaction among Shanyao, Fuling, and Yiyiren at various
proportions using response surface methodology (RSM) and the
combination index. On the basis of synergism and optimization of the
decoction process, we simultaneously conducted network pharmacology and
molecular docking experiments to elucidate the mechanisms of the
Shanyao, Fuling, and Yiyiren compound (SFY) binding with AKT1, GSK3B,
TP53, HIF1A, and PTGS2‐related targets. Then whether SFY can mitigate
oxidative stress via the AKT1/GSK3β/HIF1α antioxidant system was
evaluated, aiming to provide insight into the intervention of
H[2]O[2]‐induced oxidative injuries. Finally, a decoction‐based broad
spectrum of edible inks was selected and developed for a 3D printing
ink in a balanced diet. Collectively, the present study provides a
molecular basis for using SFY as a promising antioxidant in the future.
2. Materials and Methods
2.1. Reagents and Materials
Shanyao, Fuling, and Yiyiren were obtained from a local market of
Yangzhou. 2,2′‐Azino‐bis (3‐ethylbenzothiazoline‐6‐sulphonic acid)
(ABTS) and 2,4,6‐tris (2‐pyridyl)‐1,3,5‐triazine (TPTZ) were made in
Aldrich‐Sigma (St. Louis, MO, USA). H[2]O[2] (30%), chloroform,
anhydrous ethanol, and isopropanol were bought from Sinopharm Chemical
Reagent Co. Ltd. (China). TRIzol was obtained from Thermo Fisher
Scientific (USA). A real‐time fluorescence quantitative PCR system
(qRT‐PCR) and a reverse transcriptional kit were purchased from
TransGen Biotech (China). Superoxide dismutase (SOD) and catalase (CAT)
activity detection kits were obtained from Beyotime Biotechnology and
Grace Biotechnology (both China), respectively. Dimethyl sulfoxide
(DMSO), ethylene diamine tetraacetic acid (EDTA), phosphate buffer
solution (PBS) phosphate dry powder, and
3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) were
offered by Beijing Solarbio (China). All other agents were analytically
pure.
2.2. Optimization and Preparation of Formulations of Decoction
The preparations of medicinal and edible herbals were determined
according to the method from Guan et al. ([59]2021). Based on a
modified classic folk recipe, Sifang Tang, Shanyao, Fuling, and Yiyiren
were selected, cleaned, and air‐dried at 60°C overnight. Then the dried
materials were filtrated through an 80‐mesh sieve and kept in an
airtight container at ambient temperature until used. With the
solid/liquid ratio at 1:20, the three types of herbal powders were
randomly combined and thoroughly mixed in an RK103H ultrasonic bath
(BANDE‐LIN SONOREX, Germany) at 80 kHz. After 4 h of extraction, the
extracts were centrifuged at 4500 rpm for 10 min. The final sensory
test samples were obtained by mixing the three extracts.
2.3. Fuzzy Mathematical‐Based Sensory Evaluation
Sensory evaluation was carried out according to the procedure from Jaya
and Das ([60]2003). Briefly, 10 reportedly healthy, nonsmoking
potential evaluators from the staff and students of the College of
Tourism and Culinary Science, Yangzhou University were selected at a
60% success rate in triangle tests and enrolled to evaluate the color,
odor, taste, and texture of sensory test samples. The testing was
finished in the laboratory as per regulations in ASTM MNL‐26 (1996).
The panelists each sensorily analyzed 10 samples. The tested samples
were categorized into four distinct grades: excellent, good, fair, and
poor, which reflected their distinctiveness. The criteria for sensory
evaluation were listed in Table [61]S1, and the voting outcomes for
each criterion were recorded in Table [62]1.
TABLE 1.
Sensory evaluation indexes of polyherbal formulations.
Num. Sample Color Odor Taste Texture
V[1] V[2] V[3] V[4] V[1] V[2] V[3] V[4] V[1] V[2] V[3] V[4] V[1] V[2]
V[3] V[4]
T [1] Euryale ferox ; Lotus seed; Shanyao 4 2 4 0 3 4 2 1 3 5 1 1 2 6 2
0
T [2] Euryale ferox ; Lotus seed; Yiyiren 3 3 4 0 2 4 4 0 6 6 1 0 1 3 5
1
T [3] Euryale ferox ; Lotus seed; Fuling 0 4 3 3 0 1 4 5 1 5 3 1 5 3 2
0
T [4] Lotus seed; Shanyao; Yiyiren 5 2 2 1 4 2 2 2 2 4 4 0 3 4 2 1
T [5] Lotus seed; Shanyao; Fuling 4 5 1 0 5 3 1 1 5 5 0 0 4 4 2 0
T [6] Shanyao; Yiyiren; Fuling 6 4 0 0 7 2 1 0 7 3 0 0 3 4 1 2
T [7] Euryale ferox ; Shanyao; Yiyiren 4 4 2 0 6 3 1 0 5 4 1 0 2 8 0 0
T [8] Euryale ferox ; Shanyao; Fuling 2 1 4 3 3 4 2 1 3 7 0 0 0 6 4 0
T [9] Euryale ferox ; Yiyiren; Fuling 3 5 1 1 2 8 0 0 7 1 1 1 5 3 1 1
T [10] Lotus seed; Yiyiren; Fuling 1 6 1 2 5 3 2 0 4 4 2 0 2 5 2 1
[63]Open in a new tab
Fuzzy mathematical sensory evaluation was integrated into food sensory
assessment to quantify evaluation factors, substantially mitigate the
influence of personal bias, and produce more accurate and scientific
scoring results (Ranneh et al. [64]2021). In detail, a set of sensory
factors for the prepared solutions was defined as U, and a set of
grades as V. U comprises color (U[1]), odor (U[2]), taste (U[3]), and
texture (U[4]), and V includes excellent (V[1]), good (V[2]), average
(V[3]), and poor (V[4]). The relative weight of each sensory factor in
the overall flavor profile was determined using the fuzzy binary
comparative decision method. Ten sensory evaluators compared these
factors pairwise. A factor deemed important was given 1 point, whereas
those considered less important were assigned 0 point. The total score
of each sensory factor divided by the maximum score of 100 was used as
the weight of this sensory factor (Table [65]S2). Based on the
evaluators' scoring on the importance of color, odor, taste, and
texture in the overall sensory evaluation, the distribution of weights
among these factors is X = {color, odor, taste, texture} = {0.24, 0.28,
0.33, 0.15}.
2.4. Determination of ABTS Radical (ABTS^+) Scavenging Capacity
The ABTS^+ scavenging ability was tested following the method from
Guan, Li, et al. ([66]2024) with modifications. Briefly, an ABTS^+
stock solution was prepared by mixing 5 mL of 7 mM ABTS and 5 mL of
2.45 mM potassium persulfate and put for 12–16 h at room temperature
(RT) without light. Then the mixture was diluted using ultrapure water
until the absorbance at 405 nm was 1.4, forming an ABTS working
solution. Gradient diluted extracts or ascorbic acid (0.5 mL) were
blended with 0.5 mL of the ABTS working solution and stood for 30 min
in the dark at RT. Then the absorbance at 734 nm was recorded. The
percent of ABTS^+ scavenging activity of the extracts was computed
using Equation ([67]1):
[MATH: ABTS%=1−A1−A2/A0×100% :MATH]
(1)
where A [1], A [2], and A [0] are the absorbance of the mixture of
ABTS^+ and the sample solution, the mixture of ABTS^+ and the control
sample, and the mixture of ABTS^+ and deionized water, respectively.
2.5. Experiment Design and Mixture Optimization
Mixture design, a special type of RSM, was used to optimize the
composition of herbal mixtures and test the interactive effect between
components. Based on Section [68]2.3, the polyherbal formulations for
further optimization consisted of Shanyao, Fuling, and Yiyiren. In
total, 16 formulations were formed and the responses were analyzed
using DesignExpert 13 (Minitab Inc., State College, PA, USA). The
layout of the herbal formulations is shown in Table [69]2. The
dependent variable was in vitro ABTS antioxidant ability. The canonical
model of was used for each response after adjustment based on the
testing data. Linear, quadratic, and special cubic models were tested
to determine regression coefficients, which were kept only at the
significant level. Data were refitted to obtain the final model for
each index. The adequacy and goodness of fitting for each model were
statistically analyzed and fitted to a second‐order polynomial
regression model involving the coefficients of linear, quadratic, and
interactive terms. For validation, the optimal formulation was examined
in triplicate and expressed as mean ± standard deviation. Analysis of
variance (ANOVA) was conducted to calculate the model significance and
suitability of the factors and interactions. The formulations with
highly desirable functions were chosen for further analysis.
TABLE 2.
Mixture design experimental arrangement and results.
Standard order Factor A (Fuling, g) Factor B (Shanyao, g) Factor C
(Yiyiren, g) Response
ABTS^+ radical scavenging capacity (%)
1 0.75 0.55 0.40 41.30 ± 0.73
2 0.65 0.75 0.30 56.98 ± 0.41
3 0.65 0.65 0.40 54.35 ± 0.98
4 0.75 0.62 0.33 50.76 ± 0.47
5 0.65 0.65 0.40 61.18 ± 0.63
6 0.75 0.65 0.30 55.61 ± 0.24
7 0.65 0.75 0.30 53.13 ± 0.30
8 0.72 0.62 0.37 61.32 ± 1.09
9 0.65 0.75 0.30 54.63 ± 0.38
10 0.75 0.75 0.20 46.27 ± 0.29
11 0.55 0.75 0.40 61.15 ± 0.66
12 0.62 0.72 0.37 67.47 ± 0.68
13 0.55 0.75 0.40 57.46 ± 0.34
14 0.75 0.68 0.27 53.50 ± 0.64
15 0.68 0.68 0.33 67.60 ± 0.60
16 0.65 0.65 0.40 54.94 ± 0.21
[70]Open in a new tab
2.6. Determination of Antioxidant Synergism
Based on the Chou–Talalay combined drug theory, the synergistic
antioxidant effects of different combinations were explored
(Chou [71]2018). Specifically, antioxidant synergism was determined as
per the sum of IC[50] from seven concentrations based on RSM to
generate a total of 16 solution combinations. The absorbance of the
solution combination was then detected to determine the percent of
ABTS^+ scavenging capacity. To quantify the synergistic, additive, or
antagonistic impact of the combinations of decoction, the testing data
were converted to the combination index (CI):
[MATH: CI=MCa/SCa+MCb/SCb :MATH]
where MCa and MCb are the concentrations of compounds A and B in the
mixture to achieve 50% antioxidant activity respectively; SCa and SCb
are the EC[50] of the single compounds A and B, respectively. CI < 1,
= 1, or > 1 implies a synergistic, additive, or antagonistic effect,
respectively.
2.7. Identification of Core Ingredients, Potential Targets of Decoction, and
Network Construction
The active constituents and related targets in SFY‐based polyherbal
formulations were cited from Chinese Medicine System Pharmacology
Database ([72]https://old.tcmsp‐e.com/tcmsp.php) with pharmacokinetic
parameters (oral bioavailability [OB] ≥ 30% and drug‐like activity
[DL] ≥ 0.18). Then the target names of the corresponding core active
ingredients were turned into a unified format with Homo sapiens via
Unified Protein Database ([73]https://www.uniprot.org/). In addition,
the potential antioxidant associated targets were gathered from
GeneCards ([74]http://www.genecards.org/), Herb 2.0
([75]http://www.disgenet.org/search), and OMIM ([76]https://omim.org/).
Then duplicates were removed after the targets were merged from the
databases above. The targets of antioxidants and SFY‐based polyherbal
formulations and their common targets were acquired via Venny 2.1.0. To
systematically examine the network, the “herb‐compound‐common target”
network was imported and constructed via Cytoscape 3.9.1.
2.8. Protein–Protein Interaction (PPI) Network Construction, Topology
Analysis, and GO/KEGG Enrichment Analysis of Core Targets
To investigate the core targets and their interactions against oxidant
damages, the overlapping targets between SFY‐based polyherbal
formulations and oxidant damages were examined using the STRING network
platform ([77]https://stringdb.org/). Furthermore, a PPI network was
created at a confidence level > 0.4 and with the “ Homo sapiens ”
filter (Ren et al. [78]2024). The network was further analyzed in
Cytoscape 3.9.1 using MCODE to identify the associations between hub
genes and network clusters with specific parameters. Core clusters were
then identified based on degree centrality. GO and KEGG pathway
enrichment analyses of the core targets were done using David 6.8
([79]https://david.ncifcrf.gov/) to explore specific mechanisms and
signaling pathways. Enrichment results with p and q < 0.05 were
visualized using bar and bubble charts on the online tool imageGP.
2.9. Molecular Docking
Molecular docking, an increasingly important computational tool for
exploring the behaviors of biomacromolecule complexes, was conducted as
per the approach of Guan et al. ([80]2023) with modifications. Based on
the degrees from Section [81]2.8, the top targets and core ingredients
from the SFY‐based polyherbal formulations were chosen for affinity
calculations. The 3D structures of AKT1, GSK3B, TP53, HIF1A, and PTGS2
were cited and constructed via AutoDockTools 1.5.6. Polar hydrogens
were added to demand charges after water was removed. The smallest
binding energy was computed on AutoDock Vina. Finally, the binding
details were illustrated on PyMOL and Discovery Studio 2018.
2.10. Cytotoxicity Evaluation and Antioxidant Enzyme Activity Analysis
The BNLCL.2 mouse embryonic liver cell line from Meilun Biotechnology
Co. Ltd. (Dalian, China) was cultured in high‐glucose DMEM added with
10% FBS and 1% P/S at 37°C with 5% CO[2]. Trypsin was used for cell
passage. For H[2]O[2] cytotoxicity assessment, the BNLCL.2 cells
(1 × 10^4 cells/well) were planted in 96‐well plates and processed with
0, 200, 400, 600, 800, or 1000 μmol/L H[2]O[2] for 24 h. To evaluate
the protective effect of SFY, the cells were pretreated with 600 μM
H[2]O[2] for 4 h, and co‐cultured with 0.5, 1, 2, 4, or 8 mg/mL SFY for
24 h. In MTT assays, each well was added with 100 μL of 5 mg/mL MTT,
incubated for 4 h, and then the formazan crystals were solubilized with
DMSO. The absorbance at 490 nm was recorded using a microplate reader
(Flefel et al. [82]2019). Additionally, BNLCL.2 cells were pretreated
with H[2]O[2] or SFY at specific concentrations. Then CAT and SOD
activities were assessed using commercial kits (Cat. no. S0101S,
G0105W48) as instructed by the manufacturer, and were detected with
colorimetry and the xanthine oxidase method, respectively (Ding
et al. [83]2020).
2.11. RNA Extraction and qRT‐PCR Analysis
The BNLCL.2 cells were incubated in 6‐well plates for 24 h, pretreated
with 600 μM H[2]O[2] for 4 h, and exposed to SFY seeds (0, 2, 4, and
8 mg/mL) for 24 h. The treated BNLCL.2 cells were collected. Total RNA
was extracted from the BNLCL.2 cells using a TRIzol reagent, and
reverse transcribed to cDNA using a first‐strand cDNA synthesis
SuperMix kit for qPCR (gDNA digester plus) from TransGen Biotech
(China). Real‐time PCR was conducted with fast SYBR green master mix
(Wang et al. [84]2023). The quantities of transcripts were standardized
to that of GAPDH. The primers (Sangon Biotech, China) were presented in
Table [85]S3.
2.12. 3D Printing Ink Preparation and Printability Assessment
3D printing inks were prepared and assessed in terms of printability by
forming different structures and printing them using an in‐house 3D
printing system. The SFY decoction‐based 3D‐printed gels were prepared
following a modified version of a previous method. Specifically, 2% and
2.7% high acyl gellan gum (HAG) and gelatin (GL) hydrogels were made by
dissolving the powder in a solution at about 50°C under continuous
stirring. These hydrocolloids were then mixed with 1%, 2%, 3%, or 4%
glycerin to form the final testing gels. An extrusion‐based 3D printer
(Luckybot One, Wiiboox Technology Co. Ltd., Nanjing, China) with a
syringe and a retractable plunger was operated at around 26°C. A
0.25 mm inner diameter nozzle was used for printing, and images were
captured from the top side on a food‐grade glass plate using a mobile
phone. Uniform lighting was achieved with a white mini‐studio light box
(Rui Teng Digital, Zhejiang, China) containing 144 LED aerial lights
above the inks.
2.13. Statistical Analysis
All assays were conducted in biological triplicate unless specified
otherwise. The means of two independent groups were compared with an
unpaired two‐tailed Student's t‐test. The normally distributed data
among more than two groups were compared via one‐way ANOVA with Tukey's
post hoc test. Data were expressed as mean ± standard deviation (SD)
and analyzed on GraphPad Prism 9.5.0 (San Diego, CA, USA). The
significance levels were *p < 0.05, **p < 0.01, and ***p < 0.001.
3. Results and Discussion
3.1. Sensory and Fuzzy Comprehensive Evaluation of Optimized Compatibility
Components
To enhance the medicinal value and reduce production costs of Sifang
Tang, sensory evaluation and fuzzy mathematical methods were utilized
to optimize the active ingredients and compositions. The compatibility
of the five components (Qianshi, Shanyao, Fuling, Lianzi, and Yiyiren)
in Sifang Tang was analyzed through sensory evaluation (Table [86]1)
and radar plotting (Figure [87]1A–D). In detail, after the individual
sensory attribute for each polyherbal formulation was calculated, the
highest color was found on sample 6 (V[1] = 6), followed by sample 4
(V[1] = 5), and samples 1, 5, and 7 (V[1] = 4). As for odor, the
highest polyherbal formulation on the response scale was found in
sample 6, which linguistically means “Good.” The highest SI on taste,
flavor, and mouthfeel was found all on the response scale F4, which
linguistically implies “Good.” The SI of color was on the response
scale F3, which linguistically refers to “Satisfactory.” For sample 6,
the frequency of odor under “Excellent,” “Good,” “Average,” and “Poor”
is 7, 2, 1, and 0, and the frequency of odor is 7, 3, 0, and 0
respectively. From the above ranking, in the EFSSE fortified bread
sample, sample 6 (Shanyao, Fuling, and Yiyiren) showed better sensory
evaluation compatibility.
FIGURE 1.
FIGURE 1
[88]Open in a new tab
Optimized compound formula compatibility: sensory evaluation (A–D),
antioxidant effect (E–G), and component quantification analysis (H–J).
Sums of sensory scores for quality attributes of tested samples: T [1]
( Euryale ferox , Lotus seed, Shanyao), T [2] ( Euryale ferox , Lotus
seed, Yiyiren), T [3] ( Euryale ferox , Lotus seed, Fuling), T [4]
(Lotus seed, Shanyao, Yiyiren), T [5] (Lotus seed, Shanyao, Fuling), T
[6] (Shanyao, Yiyiren, Fuling), T [7] ( Euryale ferox , Shanyao,
Yiyiren), T [8] ( Euryale ferox , Shanyao, Fuling), T [9] ( Euryale
ferox , Yiyiren, Fuling), T [10] (Lotus seed, Yiyiren, Fuling).
Traditional sensory evaluation methods are frequently hard to achieve
consensus, owing to personal subjective perceptions, environmental
variations, and psychological fluctuations. In comparison, membership
function theory‐based fuzzy mathematical sensory evaluation can enhance
the reliability of sensory evaluations, and thus has been innovatively
integrated into food sensory assessment (Pallavi [89]2025). The sensory
evaluation data from the 10 evaluators (Table [90]1) were aggregated
into matrix R by dividing the number of votes received for each grade
by 10. The weight set X was then integrated with matrix R to form an
evaluation matrix Y = X × R. Then matrix Y was processed using a
comprehensive scoring matrix T. The set of evaluation grades K = {90,
70, 50, 30} was utilized, and each grade was multiplied by its
corresponding weight and summed to compute the total fuzzy
comprehensive evaluation score for each sample. For instance, when the
colors of medicinal solutions in the first group were evaluated: four
evaluators rated it as excellent, two as good, four4 as average, and
none as poor, resulting in R [color] = (0.4, 0.2, 0.4, 0), R
[odor] = (0.3, 0.4, 0.2, 0.1), R [taste] = (0.3, 0.5, 0.1, 0.1), and R
[texture] = (0.2, 0.6, 0.2, 0). Then we have
[MATH: Y1=R1×X=0.40.30.20.40.4
0.200.1
0.30.2<
/mtable>0.50.60.10.2<
mtd>0.10×0.24,0.28,0.33,0.15=0.309,0.415,0.21
5,0.061 :MATH]
Similarly, we have Y [2] = (0.341, 0.33, 0.427, 0.015), Y [3] = (0.108,
0.446, 0.313, 0.133), Y [4] = (0.343, 0.296, 0.266, 0.095), Y
[5] = (0.461, 0.429, 0.082, 0.028), Y [6] = (0.616, 0.311, 0.043,
0.03), Y [7] = (0.459, 0.432, 0.109, 0), Y [8] = (0.231, 0.457, 0.212,
0.1), Y [9] = (0.434, 0.422, 0.072, 0.072), and Y [10] = (0.326, 0.435,
0.176, 0.072). Then, we get T
[1] = 0.309 × 90 + 0.415 × 70 + 0.215 × 50 + 0.061 × 30 = 69.44, T
[2] = 75.59, T [3] = 60.58, T [4] = 65.74, T [5] = 72.41, T
[6] = 80.26, T [7] = 77, T [8] = 66.38, T [9] = 74.36, and T
[10] = 70.75.
With sensory evaluation of fuzzy mathematical models considered, the
polyherbal formulations can be ranked as follows: T [6] > T [7] > T
[2] > T [9] > T [5] > T [10] > T [1] > T [8] > T [4] > T [3]. After
quantitative analysis of sensory evaluation using fuzzy mathematics,
the group members/consumers believed sample T [6] (Shanyao, Fuling, and
Yiyiren) had higher overall sensory acceptability than the average.
Moreover, the overall acceptability parameter was mainly dependent on
quantitative analysis of personalized sensory attributes of the
samples, including taste, flavor, and texture. These findings indicate
the sensory characteristics of the SFY decoction are closely linked to
the proportions of Shanyao, Fuling, and Yiyiren, which could be
explained by the different molecular profiles of each herb that
contribute uniquely to taste and texture. However, the sample scope of
the current study was limited to school‐based participants and thus may
not fully reflect broader market preferences. Future work will focus on