Interval type-2 fuzzy logic system for remote vital signs monitoring and shock level prediction

Corresponding Author: uduakumoh@uniuyo.edu.ng10.22105/JFEA.2021.281500.1061
hhttp://dx.doi.org/10.22105/jfea.2021.255064.1028
E-ISSN: 2717-3453 | P-ISSN: 2783-1442
|
Abstract
1 | Introduction
Non-linear problems, especially problems that are more relevant to everyday life activities cannot
be solved with conventional models as they are not well suited. The increase in popularity of the
fuzzy logic systems in problem solving can be attributed to its ability to incorporate human
reasoning in its algorithm. The notion of Fuzzy Sets (FSs) was introduced in [1] as a method of
representing uncertainty and vagueness in a way that elements are not limited to binary
Membership Functions (MFs) of 0 or 1 but rather is a continuum in 0 and 1. A Type-1 Fuzzy
System (T1FLS) is a rule based system which can be viewed as a process that maps crisp inputs
to outputs by using the theory of fuzzy sets [2].
Journal of Fuzzy Extension and Applications
www.journal-fea.com
J. Fuzzy. Ext. Appl. Vol. 2, No. 1 (2021) 41–68.
Paper Type: Research Paper
Interval Type-2 Fuzzy Logic System for Remote Vital Signs
Monitoring and Shock Level Prediction
Uduak Umoh1,*, Samuel Udoh1, Abdultaofeek Abayomi2, Alimot Abdulazeez1
1 Department of Computer Science, University of Uyo, Uyo, PMB 1017, Uyo, Akwa Ibom State, Nigeria;
uduakumoh@uniuyo.edu.ng;samueludoh@uniuyo.edu.ng; alimotolaide4real@gmail.com.
2 Department of Information and Communication Technology, Mangosuthu University of Technology, P.O. Box 12363 Jacobs, 4026
Durban, South Africa; taofeekab@yahoo.com.
Citation:
Umoh, U., Udoh, S., Abayomi, A., & Abdulazeez, A. (2021). Interval type-2 fuzzy logic system for
remote vital signs monitoring and shock level prediction. Journal of fuzzy extension and application, 2
(1), 41-68.
Accept: 17/02/2021
Revised: 06/01/2021
Reviewed: 24/12/2020
Received: 08/11/2020
Interval Type-2 Fuzzy Logic Systems (IT2 FLSs) have shown popularity, superiority, and more accuracy in performance
in a number of applications in the last decade. This is due to its ability to cope with uncertainty and precisions adequately
when compared with its type-1 counterpart. In this paper, an Interval Type-2 Fuzzy Logic System (IT2FLS) is employed
for remote vital signs monitoring and predicting of shock level in cardiac patients. Also, the conventional, Type-1 Fuzzy
Logic System (T1FLS) is applied to the prediction problems for comparison purpose. The cardiac patients’ health datasets
were used to perform empirical comparison on the developed system. The result of study indicated that IT2FLS could
coped with more information and handled more uncertainties in health data than T1FLS. The statistical evaluation using
performance metrices indicated a minimal error with IT2FLS compared to its counterpart, T1FLS. It was generally
observed that the shock level prediction experiment for cardiac patients showed the superiority of IT2FLS paradigm
over T1FLS.
Keywords: Uncertainty, Mamdani fuzzy inference, Cardiac patient, Health data, Deffuzification.
Licensee Journal
of Fuzzy Extension and
Applications. This rticle
is an open access article
distributed under the
terms and conditions of
the Creative Commons
Attribution (CC BY)
license
(http://creativecommons.
org/licenses/by/4.0).

42
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
T1FLSs have the ability to process information using linguistic variables and make decision with
imprecise, vague, ambiguous and uncertain data. The beauty of T1FLSs is its’ ability to represent
nonlinear input/output relationships by using IF/THEN statements, called rules [3]. T1FLS consists
of four components: fuzzifier, fuzzy rules, inference engine, and defuzzifier [4]. T1FLSs have
achieved great success in handling many different real-world problems such as classification,
regression, control, decision making, prediction, and so on [5-11]. However, due to the complexity
and uncertainty in many real-world problems, T1FLs cannot adequately cope with or minimize the
effects of the uncertainties posed by the complex nature of many real-world problems. These
uncertainties can be as a result of uncertainty in inputs, uncertainty in outputs, uncertainty that is
related to the linguistic differences, uncertainty caused by the change of conditions in the operation
and uncertainty associated with the noisy data when training the fuzzy logic system [12].
In order to address this problem, [13] has recognized this potential limitation and introduced a higher
type of fuzzy sets which is the concept of Type-2 Fuzzy Logic System (T2FLS) from Type-2 Fuzzy
Set (T2FS). T2FLS is an extension of T1FLS with additional design degrees of freedom where the
MFs are themselves fuzzy with the actual membership grade of an element assumed to lie within a
closed interval of 0 and 1. However, T2FLS suffers computational complexity and to resolve this
problem, Interval Type-2 Fuzzy Logic System (IT2FLS) is used which is a simplified version of
T2FLS with reduced computational intensity making it quite practicable. Recently, T2FLSs and
IT2FLSs have been successfully applied to handle a great deal of uncertainties in prediction problems
and the results are very encouraging [14-15]. Consequently, IT2FLSs have dominated the research
field and T2FLS are widely applied in various are due to their simpler structure and reduced
computational cost [16]-[23].
In this study, remote vital signs monitoring and prediction of shock level in cardiac patients using an
IT2FLS based on Mamdani fuzzy inference is presented. Our motivation is to apply IT2FS to reduce
prediction error in the task of modeling uncertainties in health data. Mamdani Fuzzy Inference
System (FIS) is employed as the background algorithm. Also, the study investigates the prediction of
problems using conventional T1FLS for comparison purpose. The rest of the paper is structured as
follows: Section 2 involves the preliminaries and reviews of the concepts on TIFS and IT2IFS. In
Section 3, the design of IT2FLS for remote vital signs monitoring and prediction of shock level in
cardiac patients is carried out. Our model results and discussion are presented in Section 4 and
Section 5 involves conclusion of study.
2| An Overview of Type-1 and Interval Type-2 Fuzzy Sets
In this section, basic concepts of TIFS and IT2IFS are reviewed to deliver the fundamental
knowledge required for comprehension of the work.
2.1| Type-1 Fuzzy Sets
According to [24], a classical fuzzy set otherwise called a T1FS, denoted by A, is characterized by a
Type-1 Membership Function(T1MF), 𝜇𝐴(𝑥) where 𝑥 ∈ 𝑋, and 𝑋 is the domain of definition of the
variable as seen in Eq. (1).
A = {{(x,μA(x)}|∀x ∈ X} (1)

43
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
Where 𝜇𝐴 is called aType-1 membership function of T1FS, 𝐴 and maps each element of A to a
membership value called membership grade between 0 and 1. Eqs. (2) and (3) represent the T1FS, 𝐴 in
its discrete and continuous forms respectively.
Where 𝑥1, 𝑥2, 𝑥𝑛 are members of fuzzy set A and µ𝐴(𝑥1), µ𝐴(𝑥2), µ𝐴(𝑥𝑛) are their respective membership
grades.
2.2|Type-1 Fuzzy Logic Systems (T1FLS)
The T1FLS also known as T1FIS is both, intuitive and numerical. The inference engine combines
rules and gives a mapping from input type-1 fuzzy sets to output type-1 fuzzy sets. A fuzzy rule is a
conditional statement in the form of IF-THEN where it contains two parts, the IF part called the
antecedent part and the THEN part called the consequent part. Every T1FIS is associated with a set
of rules obtained either from numerical data, or from problem domain experts with meaningful
linguistic interpretations. In literature, we have two rules methods: Mamdani fuzzy rules [25] and [26]
and Tagaki Sugeno Kang (TSK) [27] rules respectively. In Mamdani rule, the rule consequents are
fuzzy sets while in TSK, the rule consequents are crisp functions of the inputs and can be illustrated
in the form as follows:
Where, y is a T1FS, and a, b, c are crisp coefficients.
The membership grades in the fuzzy input sets are combined with those in the fuzzy output sets using
the composition method and the outcome is combined with rules in an antecedent/consequent
format, and then aggregated according to approximate reasoning theory. The structure of a T1FIS is
shown in Fig. 1 and consists of four basic components namely: the fuzzifier, the fuzzy rule-base, the
inference engine, and the defuzzifier.
Fig. 1. Structure of a type 1 fuzzy logic system [18].
A = µA(x1)/ x1 + µA(x2)/ x2 + ⋯ + µA(xn)/ xn, (2)
A = ∫ µA(x)/x
X
. (3)
Mamdani: RkIF xi is A i
k
and,., and xn is A n
k
then yk is Bi
k
. (4)
TSK: RkIF xi is A i
k
and,., and xn is A n
k
then yk = a ∗ x1 + b ∗ x2 + c. (5)

44
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
The fuzzifier maps the crisp input into a T1FS sets by evaluating the crisp inputs based on the
antecedent part of the rules and assigns each crisp input a degree of membership in its input fuzzy
set. The fuzzy rule-base is a collection of IF–THEN statements called rules in the form represented
in Eqs. (4) and (5). The inference engine combines the type-1 fuzzy input set and the rules to produce
a type-1 fuzzy output and the defuzzifier maps the type-1 fuzzy output set, T1FSs that is produced
by the inference engine into crisp values. The centroid of area defuzzification method based on [24]
is applied to compute the crisp number as seen in Eq. (6).
Where, ∑ 𝑌
́𝑘𝑍𝑘
𝑛
𝑖=1 is the membership value in the membership function and 𝑌
́𝑘is the centre of
membership function which is the running point in a discrete universe. The expression can be
interpreted as the weighted average of the elements in the support set [28].
2.3|Interval Type-2 Fuzzy Set (IT2FS)
An IT2FS is a simplified version of T2FS (1) introduced in [13] as an extension of T1FSand is
characterized by a Type-2 Membership Function (T2MF), 𝜇Ã(𝑥, 𝑢) and 0 ≤ 𝜇Ã(𝑥,𝑢) ≤ 1 as represented
in Eq. (7) [12]. For an IT2FS, for which X and U are discrete, the domain of Ã is equal to the union
of all of its embedded T1 FSs as seen in Eq. (8).
Where 𝑥 ∈ 𝑋 and 𝑢 ∈J ⊆ [0, 1x]; x is the primary variable with a domain X and u∈U is the secondary
variable with domain 𝐽𝑥 at each x∈X. Jx is the primary membership of x and the secondary grades of
all equal 1 [15, 18]. Hence, a type-2 membership grade can be any subset in [0, 1] [12]. Uncertainty
about Ã is conveyed by the union of all the primary memberships, which is called the Footprint Of
Uncertainty (FOU) of A as shown in Eq. (9) and Fig. 2, respectively.
Fig. 2. Interval type-2 fuzzy set [29].
Crisp Output, Z = ∑ Ý kZk
n
i =
∑ Ý k𝐙𝐤
𝐧
𝐢=𝟏
∑ Ý 𝐤
𝐧
𝐢=𝟏
=
∑ Zk
𝐧
𝐢=𝟏
∑ 𝐘𝐤
𝐧
𝐢=𝟏
. (6)
Ã = {((x,u), μÃ(x, u))|∀ x ∈ X, ∀ u ∈Jx ⊆ [0, 1], (7)
Ã = ∑ [∑ [1/u]
u∈Jx ]
n
i=1 /xi . (8)
μÃ(x, u) = 1, FOU (Ã) = ⋃ Jx
∀x∈X
= {(x, u): u ∈ Jx ⊆ [0, 1]}. (9)

45
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
The Upper Membership Function (UMF) and Lower Membership Function (LMF) of Ãcomprises
two type-1 MFs that bound the FOU. The UMF is associated with the upper bound of FOU (Ã) and
is denoted by 𝜇
Ã
(𝑥),∀x∈X, and the LMF is associated with the lower bound of FOU (Ã) and is denoted
𝜇Ã(x), ∀x∈X, that is:
From Eq. (3),
The LMF of Ã are twice T1MFs bound the FOU in Eqs.(9) and (10) and 𝐽𝑥 is an interval set. The set
theory operations of union, intersection and complement are applied to compute IT2FSs. For an
IT2FS, for which X and U are discrete, the domain of Ã is equal to the union of all of its embedded
T1 FSs, so that Ã can be expressed as;
3|Interval Type-2 Fuzzy Logic System (IT2FLS)
Fig. 3 shows a structure of an IT2FLS which is a version of T2FL. IT2FLS is a FLS that uses at least
one IT2FS in mapping from crisp inputs to crisp outputs. IT2FLS is employed to reduce the
computational burden of T2FLS while preserving the advantages of T2FLS. T1FLS cannot adequately
handle the effect of uncertainties posed by the complex nature of real-world data because it uses T1FS
which is characterized by membership functions that are certain. To address this problem, IT2FL is
introduced as an extension of T1FL to provide additional design degrees of freedom with membership
functions that are themselves fuzzy. IT2FL can be very useful when used in situations where lots of
uncertainties are presented and have the potential to provide better performance than T1FLS. An
IT2FIS is characterized by five components: a fuzzifier, a rule-base, an inference-engine, type-reducer
and defuzzifier - that are inter-connected. Also, an IT2FIS is characterized by IF–THEN rules, in this
case, the antecedent and consequents parts are Type-2 fuzzy sets [12].
Fig. 3. The structure of a type -2 fuzzy logic system [30].
μÃ
(x) ≡ FOU (Ã) ∀x ∈ X, (10)
μÃ(x) ≡ FOU (Ã) ∀x ∈ X, (11)
Jx = {(x, u): u ∈ [μÃ(x), μÃ
(x)]}. (12)
Ã = FOU (Ã) = ⋃ [μÃ(x) ,
∀x∈X
μ
Ã
(x)]. (13)
Ã = 1/FOU (Ã) = 1/ ⋃[
x∈X
μÃ(x), μÃ
(x)]. (14)

46
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
Where 𝑥𝑖 , i = 1, 2…, n are the antecedents, 𝑦 is the consequent of the kth rule of IT2FLS. The 𝐴̃𝑖
’s are
the MFs 𝜇𝐴̃
𝑖
𝑘(𝑥𝑖) of antecedent part of the ith input 𝑥𝑖 , The 𝐵𝑖
is the MFs 𝜇 𝐵̃
𝑖
𝑘(𝑦) of the consequent
part of the output 𝑦𝑗. The firing strength is evaluated from the input and antecedent operations to
produce an IT1 set [31].
Where 𝐹𝑖
(𝑥′
) is the antecedent of rule i and µF1
i
(xˈ) is the degree of membership of x in F.𝜇𝑓̃𝑖(x) and
𝜇𝑓̃𝑖(x) are upper and lower MFs of 𝜇𝑓̃𝑖, i = 1 to m respectively. The inference engine combines the
fired rules to produce a mapping from input to output in IT2FSs. The output fuzzy set, 𝜇
𝐸̃
𝑗
𝑙(𝑦𝑗), is
evaluated by combining the fired output consequent sets through the union of the kth rule fired
output consequent sets. An exact iterative method of type-reduction is performed to compute the
centroid of an IT2FS which is a T1FS. The type-reduced set gives an interval of uncertainty for the
output of an IT2FLS: the more uncertainties in IT2FLS, the more uncertainties about its MFs and
the larger the type-reduced set, and vice-versa. IT2FS are characterized by their left- and right-end
points required to compute the centroid of an IT2FS [12], [32] and a detailed definition oftype-
reduction Eq. (17) can be found in [15] and [30], [31] respectively. In this paper, we adopt Karnik
and Mendel (KM), Algorithms [32] to calculate the exact end-points in Eqs.(18) and (19), respectively
we obtain the defuzzified crisp output for each output k using Eq. (20).
The
performance criteria in Eqs. (21) and (22) are defined and applied to measure our experimental
results: Mean Squared Error (MSE) and Root Mean Squared Error (RMSE).
Mamdani: RkIF xi is A
̃ i
k
and, … , and xn is A
̃ n
k
then yk is B̃i
k
. (15)
Fi
(x′
) = [f′
(x′
),f′(x′
)] ≡ [f′
, f′ (16)
YTR(x′
) = [yl(x′
), yr(x′
)] ≡ [yl, yr]
= ∫ … ∫ ∫ . . ∫ 1/
∑ fi
y i
N
i=1
∑ fi
N
i=1
fN∈[fN,f
N
]
f1∈[f1,f
1
]
y1∈[yl
N
,yr
N]
y1∈[yl
1
,yr
1]
. (17)
yr =
∑ fr
i yr
i
N
i=1
∑ fr
i
N
i=1
. (18)
𝑦𝑙 =
∑ 𝑓𝑙
𝑖
𝑦𝑙
𝑖
𝑁
𝑖=1
∑ 𝑓𝑙
𝑖
𝑁
𝑖=1
. (19)
Yk(X) =
ylk+yrk
2
. (20)
MSE =
1
N
∑(y x
− y)2
,
N
i=1
(21)
RMSE = √
1
N
∑(yx − y)2
N
i=1
. (22)

47
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
Where 𝑦𝑥
is desired output, y is the computed output and N is the number of data items, respectively.
4|Model Experiment
The model experiment was carried out in two (2) major stages namely: clinical data collection and
design of interval type-2 fuzzy logic model for prediction problems in patients. Data of cardiovascular
patients were collected. The data were pre-processed and stored in a database. The interval type-2
fuzzy logic model was designed. Inputs to the model were systolic blood pressure, diastolic blood
pressure, temperature, heart rate and respiratory rate while cardiac shock level served as the desired
output. The model employed, triangular membership function, Mamdani inference strategy, Karnik-
Mendel reduction method and centroid deffuzification process for cardiac shock prediction in
patients.
4.1| Clinical Data Collection
Clinical data were collected for cardiovascular patients at the Federal Medical Centre Yenagoa, Bayelsa,
Nigeria and University of Uyo Teaching Hospital, Akwa Ibom State, Nigeria, respectively for the
period of 2016-2018. To represent ethical concerns, a written permission was processed and duly
approved before data collection commenced. Expert knowledge on cardiovascular disease and disease
diagnosis was captured with 50 questionnaire and interview from the head of cardiology unit in the
department of medicine in both Federal Medical Centre, Yenagoa, Bayelsa State, Nigeria and
University of Uyo Teaching Hospital, Akwa Ibom State, Nigeria, respectively. A total of 1000 patients
sample data were obtained from the University of Uyo Teaching Hospital, and in the Federal Medical
Centre Yenagoa, Bayelsa, Nigeria. The sample dataset for the first 35 patients extracted from Federal
Medical Centre Yenagoa are presented in the Table 1.
4.2| Interval Type-2 Fuzzy Logic Model for Prediction Problems in Patients
The data collected in Section 4.1 were stored in a database and served as input to the interval type-2
fuzzy logic controller. In the fuzzy controller system, the data were fuzzified using fuzzy linguistic
variable and membership triangular membership function. Inference was made based on fuzzy rules
which produced interval type 2 fuzzy set. Karnik-Mendel method was employed to reduce the type 2
fuzzy set to type 1 fuzzy set. The fuzzy output was obtained through centroid defuziffication process.
Notification message module was incorporated in the system. This would enable alerts and vital
information to be sent to doctors-on-call, caregivers, ambulance agency and designated family
members for necessary action based on detection threshold from the cardiac shock prediction results.
Interval Type-2 Fuzzy logic frameworks for prediction problems in cardiac patient is presented in Fig.
4.

48
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
Fig. 4. Interval type-2 fuzzy logic framework for prediction problems in cardiac patient.
Table 1. Sample data extraction from Federal Medical Centre Yenagoa, Bayelsa
Age Sex BPDiastolic BPSystolic Temp oC Weight (kg) Respiratory
Rate(cm)
Pulse
Rate(b/m)
70 F 120 90 36.9 82 20 80
53 F 140 100 36.7 90 22 68
60 M 170 100 36.1 77.4 24 120
45 M 100 70 35.2 80 22 70
57 M 105 70 36.9 79 18 122
63 F 170 90 36.2 77.4 20 80
37 M 125 80 36.4 62.5 20 80
41 F 120 70 35.2 55 20 86
56 M 130 90 36.8 92 20 80
60 F 150 110 36.9 77.4 20 86
57 M 98 60 36.6 91 20 76
56 M 120 80 36.2 79 22 73
44 M 140 90 36.9 55 20 80
52 M 120 90 36.7 48 20 90
57 F 150 120 39 50 22 80
54 M 120 80 36.8 80 24 78
48 M 112 60 36.7 50 24 62
49 F 110 80 36.3 41.4 20 60
64 F 140 80 36.7 94.1 16 78
58 F 120 70 36.7 80 20 82
50 F 130 80 36.4 49.8 20 68
59 M 130 100 36.1 53 22 90
66 F 150 90 36.4 93 21 78
43 M 130 80 36.5 50.6 23 64
69 M 160 90 36.7 96.5 20 80
59 M 130 90 36.4 51.2 19 68
44 M 120 80 36.7 68 17 59
42 M 140 80 36.7 71 20 72
61 F 145 100 36.8 58 22 85
40 M 90 70 35.1 62 24 69
71 M 120 70 36.3 50 20 70
59 F 120 80 36.4 70 22 90
51 M 120 80 37.4 60 21 80
65 F 130 80 36.3 68.8 20 68
53 M 190 110 36.1 59.5 24 78

49
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
In this paper, derived input parameters are defined as temperature, heart rate, blood pressure and
respiratory rate, the membership functions and the associated linguistic variables, the universe of
discourse and fuzzy linguistic variable are well defined. Table 2 presents the input variables and the
universe of discourse.
Table 2. The input variables and the universe of discourse.
Triangular Membership Function (TMF) method in Eq. (23) was used to evaluate each input and
output Membership Functions (MFs) as follows:
Where, a, b, c represent the x coordinates of the three vertices of 𝜇𝐴(𝑥)in a fuzzy set A. a , c is the
lower boundary and Upper boundary where membership degree is zero respectively, b is the centre
where membership degree is 1 . x is the coordinate of three vertices (taking by assumption). The
IT2FL UMF and LMF are derived from general model in Eqs.(24) and (25) and evaluated for all the
input variables respectively. The Membership function values for shock prediction are shown in Table
3.
Where y is left end point of both UMF and LMF and r is right end point of both UMF and LMF and
q is the peak point.
Input Variables and their Universe of Discourse
Heart
Rate
Respiratory
Rate
Body
Temperature
Blood Pressure Prediction
Systolic Diastolic
[0, 200] [0, 60] [0 ,41.5] [0, 200] [0, 150] [0 , 1]
=
{
0 if x ≤ a
x−a
b−a
if a ≤ x ≤ b
c−x
c−b
if b ≤ x ≤ c
0 if x ≥ a
(
μ
̅(x)
IT2FL UMF
) =
{
0, x ≤ y1
x − y1
q1−y1
, if y1 ≤ x ≤ q1
1, if q1 ≤ x ≤ r2
r2 − x
r2 − q2
, if q2 ≤ x ≤ r2
.
(
μ(x)
IT2FL LMF
) =
{
0, x ≤ y2
x − y2
q2−y2
, if y2 ≤ x ≤
r1(q2−y2
)+y2(r1−q1
)
(q2−y2
)+(r1−q1
)
r2 − x
r2 − q2
, if x >
r1(q2−y2
)+y2(r1−q1
)
(q2−y2
)+(r1−q1
)
<x<r1
0, if x ≥ r2
.

50
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
Respiratory rate UMF and LMF (low)
Respiratory Rate UMF and LMF Normal (NM)
For Respiratory Rate UMF and LMF High (H)
or Heart Rate UMF and LMF low (L)
𝜇𝑅𝑅(𝑁)(𝑥)
=
{
0, 𝑥 ≤ 10.13
𝑥 − 10.13
𝑞2−10.13
, 𝑖𝑓 10.13 ≤ 𝑥 ≤
26.8(𝑞2−10.13)+10.13(26.8−𝑞1
)
(𝑞2−10.13)+(26.8−𝑞1
)
29.6 − 𝑥
29.6 − 𝑞2
, 𝑖𝑓 𝑥 >
26.8(𝑞2−10.13)+10.13(26.8−𝑞1
)<𝑥<26.8
(𝑞2−10.13)+(26.8−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 29.6
𝜇
𝑅𝑅
(𝑁)(𝑥)
=
{
0, ≤ 7.61
𝑥 − 7.61
𝑞1−7.61
, 𝑖𝑓 7.61 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
29.6 − 𝑥
29.6 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 29.6
0, 𝑖𝑓 𝑥 > 29.6
μRR(H)(x)
=
{
0, x ≤ 26.1
x − 26.1
q2−26.1
, if 26.1 ≤ x ≤
54.8(q2−26.1)+26.1(54.8−q1
)
(q2−26.1)+(26.8−q1
)
59.8 − x
59.8 − q2
, if x >
54.8(q2−26.1)+26.1(54.8−q1
)<x<54.8
(q2−26.1)+(54.8−q1
)
0, if x ≥ 59.8
𝜇
𝑅𝑅
(𝐻)(𝑥)
=
{
0, 𝑥 ≤ 22.8
𝑥 − 22.8
𝑞1−22.8
, 𝑖𝑓 22.8 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
59.8 − 𝑥
59.8 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 59.8
0, 𝑖𝑓 𝑥 > 59.8
𝜇
𝑅𝑅
(𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 0.0863
𝑥 − 0.0863
𝑞1−0.0863
, 𝑖𝑓 0.0863 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
13.4 − 𝑥
13.4 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 13.4
0, 𝑖𝑓 𝑥 > 13.4
𝜇𝑅𝑅(𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 1.69
𝑥 − 𝑦2
𝑞2−𝑦2
, 𝑖𝑓 1.69 ≤ 𝑥 ≤
12.01(𝑞2−1.69)+1.69(12.01−𝑞1
)
(𝑞2−1.69)+(12.01−𝑞1
)
13.4 − 𝑥
13.4 − 𝑞2
, 𝑖𝑓 𝑥 >
12.01(𝑞2−1.69)+1.69(12.01−𝑞1
)<𝑥<12.01
(𝑞2−1.69)+(12.01−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 13.4
𝜇𝐻𝑅
(𝐿)(𝑥)
=
{
0, 𝑥 ≤ −0.208
𝑥 − (−0.208)
𝑞1−(−0.208)
, 𝑖𝑓 − 0.208 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
60 − 𝑥
60 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 60
0, 𝑖𝑓 𝑥 > 60
𝜇𝐻𝑅(𝐿)(𝑥)
=
{
0, 𝑥 ≤ 4.48
𝑥 − 4.48
𝑞2−4.48
, 𝑖𝑓 4.48 ≤ 𝑥 ≤
56.2(𝑞2−4.48)+4.48(56.2−𝑞1
)
(𝑞2−4.48)+(56.2−𝑞1
)
60 − 𝑥
60 − 𝑞2
, 𝑖𝑓 𝑥 >
56.2(𝑞2−4.48)+4.48(56.2−𝑞1
)<𝑥56.2
(𝑞2−4.48)+(56.2−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 60

51
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
or Heart Rate UMF and LMF Normal (N)
or Heart Rate UMF and LMF High (H)
For Body Temperature UMF and LMF (low)
For Body Temperature UMF and LMF (Normal)
𝜇
𝐻𝑅
(𝑁)(𝑥)
=
{
0, 𝑥 ≤ 49.5
𝑥 − 49.5
𝑞1−49.5
, 𝑖𝑓 49.5 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
123 − 𝑥
123 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 123
0, 𝑖𝑓 𝑥 > 123
𝜇𝐻𝑅(𝑁)(𝑥)
=
{
0, 𝑥 ≤ 53.03
𝑥 − 53.03
𝑞2−53.03
, 𝑖𝑓 4.48 ≤ 𝑥 ≤
56.2(𝑞2−53.03)+53.03(56.2−𝑞1
)
(𝑞2−53.03)+(56.2−𝑞1
)
123 − 𝑥
123 − 𝑞2
, 𝑖𝑓 𝑥 >
119(𝑞2−53.03)+53.03(119−𝑞1
)<𝑥119
(𝑞2−53.03)+(119−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 123
𝜇
𝐻𝑅
(𝐻)(𝑥)
=
{
0, 𝑥 ≤ 107
𝑥 − 107
𝑞1−107
, 𝑖𝑓 107 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
198 − 𝑥
198 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 198
0, 𝑖𝑓 𝑥 > 198
𝜇𝐻𝑅(𝐻)(𝑥)
=
{
0, 𝑥 ≤ 111
𝑥 − 111
𝑞2−111
, 𝑖𝑓 111 ≤ 𝑥 ≤
191(𝑞2−111)+111(191−𝑞1
)
(𝑞2−111)+(191−𝑞1
)
198 − 𝑥
198 − 𝑞2
, 𝑖𝑓 𝑥 >
191(𝑞2−111)+111(191−𝑞1
)<𝑥191
(𝑞2−111)+(191−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 198
𝜇
𝑇𝑒𝑚𝑝
(𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 0.145
𝑥 − 0.145
𝑞1−0.145
, 𝑖𝑓 0.145 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
22.5 − 𝑥
22.5 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 22.5
0, 𝑖𝑓 𝑥 > 22.5
𝜇𝑇𝑒𝑚𝑝(𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 1.85
𝑥 − 1.85
𝑞2−1.85
, 𝑖𝑓 1.85 ≤ 𝑥 ≤
20.3(𝑞2−1.85)+1.85(20.3−𝑞1
)
(𝑞2−1.85)+(20.3−𝑞1
)
22.5 − 𝑥
22.5 − 𝑞2
, 𝑖𝑓 𝑥 >
20.3(𝑞2−1.85)+1.85(20.3−𝑞1
)<𝑥20.3
(𝑞2−1.85)+(20.3−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 22.5
𝜇
𝑇𝑒𝑚𝑝
(𝑛𝑜𝑟𝑚𝑎𝑙)(𝑥)
=
{
0, 𝑥 ≤ 9.59
𝑥 − 9.59
𝑞1−9.59
, 𝑖𝑓 9.59 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
30.02 − 𝑥
30.02 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 30.02
0, 𝑖𝑓 𝑥 > 30.02
𝜇𝑇𝑒𝑚𝑝(𝑛𝑜𝑟𝑚𝑎𝑙)(𝑥)
=
{
0, 𝑥 ≤ 11.48
𝑥 − 11.48
𝑞2−11.48
, 𝑖𝑓 11.48 ≤ 𝑥 ≤
28.3(𝑞2−11.48)+11.48(28.3−𝑞1
)
(𝑞2−11.48)+(28.3−𝑞1
)
30.02 − 𝑥
30.02 − 𝑞2
, 𝑖𝑓 𝑥 >
28.3(𝑞2−11.48)+11.48(28.3−𝑞1
)<𝑥28.3
(𝑞2−11.48)+(28.3−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 30.02

52
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
For Body Temperature UMF and LMF (High)
For Body Temperature UMF and LMF very high
For Systolic Blood Pressure UMF and LMF for Low
For Systolic Blood Pressure UMF and LMF (Normal)
𝜇𝑇𝑒𝑚𝑝
(ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 25.1
𝑥 − 25.1
𝑞1−25.1
, 𝑖𝑓 25.1 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
37.7 − 𝑥
37.7 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 37.7
0, 𝑖𝑓 𝑥 > 37.7
𝜇𝑇𝑒𝑚𝑝(ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 26.0
𝑥 − 26.0
𝑞2−26.0
, 𝑖𝑓 26.0 ≤ 𝑥 ≤
36.3(𝑞2−26.0)+26.0(36.3−𝑞1
)
(𝑞2−26.0)+(36.3−𝑞1
)
37.7 − 𝑥
37.7 − 𝑞2
, 𝑖𝑓 𝑥 >
36.3(𝑞2−26.0)+26.0(36.3−𝑞1
)<𝑥36.3
(𝑞2−26.0)+(36.3−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 37.7
μTemp
( veryhigh)(x)
=
{
0, x ≤ 30.5
x − 30.5
q1−30.5
, if 30.5 ≤ x ≤ q1
1, if q1 ≤ x ≤ q2
41.4 − x
41.4 − q2
, if q2 ≤ x ≤ 41.4
0, if x > 41.4
𝜇𝑇𝑒𝑚𝑝(𝑣𝑒𝑟𝑦ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 31.9
𝑥 − 31.9
𝑞2−31.9
, 𝑖𝑓 31.9 ≤ 𝑥 ≤
40(𝑞2−31.9)+31.9(40−𝑞1
)
(𝑞2−31.9)+(40−𝑞1
)
41.4 − 𝑥
41.4 − 𝑞2
, 𝑖𝑓 𝑥 >
40(𝑞2−31.9)+31.9(40−𝑞1
)<𝑥<40
(𝑞2−31.9)+(40−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 41.4
𝜇
𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃
( 𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 2.34
𝑥 − 2.34
𝑞1−2.34
, 𝑖𝑓 2.34 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
90.03 − 𝑥
90.03 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 90.03
0, 𝑖𝑓 𝑥 > 90.03
𝜇𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 11.2
𝑥 − 11.2
𝑞2−11.2
, 𝑖𝑓 11.2 ≤ 𝑥 ≤
84(𝑞2−11.2)+11.2(84−𝑞1
)
(𝑞2−11.2)+(84−𝑞1
)
90.03 − 𝑥
90.03 − 𝑞2
, 𝑖𝑓 𝑥 >
84(𝑞2−11.2)+11.2(84−𝑞1
)<𝑥<84
(𝑞2−11.2)+(84−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 90.03
𝜇𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃
( 𝑛𝑜𝑟𝑚𝑎𝑙)(𝑥)
=
{
0, 𝑥 ≤ 74.42
𝑥 − 74.42
𝑞1−74.42
, 𝑖𝑓74.42 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
139 − 𝑥
139 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 139
0, 𝑖𝑓 𝑥 > 139
𝜇𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑛𝑜𝑟𝑚𝑎𝑙)(𝑥)
=
{
0, 𝑥 ≤ 78.76
𝑥 − 78.76
𝑞2−78.76
, 𝑖𝑓 78.76 ≤ 𝑥 ≤
135(𝑞2−78.76)+78.76(135−𝑞1
)
(𝑞2−78.76)+(135−𝑞1
)
139 − 𝑥
139 − 𝑞2
, 𝑖𝑓 𝑥 >
135(𝑞2−78.76)+78.76(135−𝑞1
)<𝑥<135
(𝑞2−78.76)+(135−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 139

53
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
or Systolic Blood Pressure UMF and LMF (High)
For Systolic Blood Pressure UMF and LMF Severe
For Diastolic Blood Pressure UMF and LMF Low
( ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 151
𝑥 − 151
𝑞1−151
, 𝑖𝑓151 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
159.7 − 𝑥
159.7 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 159.7
0, 𝑖𝑓 𝑥 > 159.7
𝜇𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 128
𝑥 − 128
𝑞2−128
, 𝑖𝑓 128 ≤ 𝑥 ≤
154(𝑞2−128)+128(154−𝑞1
)
(𝑞2−128)+(154−𝑞1
)
159.7 − 𝑥
159.7 − 𝑞2
, 𝑖𝑓 𝑥 >
154(𝑞2−128)+128(154−𝑞1
)<𝑥<154
(𝑞2−128)+(154−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 159.7
( 𝑣𝑒𝑟𝑦ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 151
𝑥 − 151
𝑞1−151
, 𝑖𝑓 151 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
179.9 − 𝑥
179.9 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 179.9
0, 𝑖𝑓 𝑥 > 179.9
𝜇𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑣𝑒𝑟𝑦ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 155
𝑥 − 155
𝑞2−155
, 𝑖𝑓 155 ≤ 𝑥 ≤
177(𝑞2−155)+155(177−𝑞1
)
(𝑞2−155)+(177−𝑞1
)
179.9 − 𝑥
179.9 − 𝑞2
, 𝑖𝑓 𝑥 >
177(𝑞2−155)+155(177−𝑞1
)<𝑥<177
(𝑞2−155)+(177−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 179.9
𝜇
𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃
( 𝑠𝑒𝑣𝑒𝑟𝑒)(𝑥)
=
{
0, 𝑥 ≤ 171
𝑥 − 171
𝑞1−171
, 𝑖𝑓 171 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
198 − 𝑥
198 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 198
0, 𝑖𝑓 𝑥 > 198
𝜇𝑠𝑦𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑣𝑒𝑟𝑦ℎ𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 174
𝑥 − 174
𝑞2−174
, 𝑖𝑓174 ≤ 𝑥 ≤
194(𝑞2−174)+174(194−𝑞1
)
(𝑞2−174)+(194−𝑞1
)
198 − 𝑥
198 − 𝑞2
, 𝑖𝑓 𝑥 >
194(𝑞2−174)+174(194−𝑞1
)<𝑥<194
(𝑞2−174)+(194−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 198
𝜇𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃
( 𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 1.778
𝑥 − 1.778
𝑞1−1.778
, 𝑖𝑓 1.778 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
67.53 − 𝑥
67.53 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 67.53
0, 𝑖𝑓 𝑥 > 67.53
𝜇𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑙𝑜𝑤)(𝑥)
=
{
0, 𝑥 ≤ 6.45
𝑥 − 6.45
𝑞2−6.45
, 𝑖𝑓6.45 ≤ 𝑥 ≤
63.84(𝑞2−6.45)+6.45(63.84−𝑞1
)
(𝑞2−6.45)+(63.84−𝑞1
)
67.53 − 𝑥
67.53 − 𝑞2
, 𝑖𝑓 𝑥 >
63.84(𝑞2−6.45)+6.45(63.84−𝑞1
)<𝑥<63.84
(𝑞2−6.45)+(63.84−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 67,53

54
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
For Diastolic Blood Pressure UMF and LMF Normal
For Diastolic Blood Pressure UMF and LMF High
For Diastolic Blood Pressure UMF and LMF Very High
For Diastolic Blood Pressure UMF and LMF Severe
𝜇
𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃
(𝑁𝑜𝑟𝑚𝑎𝑙)(𝑥)
=
{
0, 𝑥 ≤ 54.29
𝑥 − 54.29
𝑞1−54.29
, 𝑖𝑓 54.29 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
102.8 − 𝑥
102.8 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 102.8
0, 𝑖𝑓 𝑥 > 102.8
𝜇𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑁𝑜𝑟𝑚𝑎𝑙)(𝑥)
=
{
0, 𝑥 ≤ 57.3
𝑥 − 57.3
𝑞2−57.3
, 𝑖𝑓57.3 ≤ 𝑥 ≤
98.5(𝑞2−57.3)+57.3(98.5−𝑞1
)
(𝑞2−57.3)+(98.5−𝑞1
)
102.8 − 𝑥
102.8 − 𝑞2
, 𝑖𝑓 𝑥 >
98.5(𝑞2−57.3)+57.3(98.5−𝑞1
)<𝑥<98.5
(𝑞2−57.3)+(98.5−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 102.8
(𝐻𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 90.6
𝑥 − 90.6
𝑞1−90.6
, 𝑖𝑓 90.6 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
117 − 𝑥
117 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 117
0, 𝑖𝑓 𝑥 > 117
𝜇𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝐻𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 92.3
𝑥 − 92.3
𝑞2−92.3
, 𝑖𝑓92.3 ≤ 𝑥 ≤
115(𝑞2−92.3)+92.3(115−𝑞1
)
(𝑞2−92.3)+(115−𝑞1
)
117 − 𝑥
117 − 𝑞2
, 𝑖𝑓 𝑥 >
115(𝑞2−92.3)+92.3(115−𝑞1
)<𝑥<115
(𝑞2−92.3)+(115−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 117
𝜇
𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃
(𝑉𝑒𝑟𝑦𝐻𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 111.3
𝑥 − 111.3
𝑞1−111.3
, 𝑖𝑓 111.3 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
134.9 − 𝑥
134.9 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 134.9
0, 𝑖𝑓 𝑥 > 134
𝜇𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑉𝑒𝑟𝑦𝐻𝑖𝑔ℎ)(𝑥)
=
{
0, 𝑥 ≤ 112.4
𝑥 − 112.4
𝑞2−112.4
, 𝑖𝑓112.4 ≤ 𝑥 ≤
134(𝑞2−112.4)+112.4(134−𝑞1
)
(𝑞2−112.4)+(134−𝑞1
)
134.9 − 𝑥
134.9 − 𝑞2
, 𝑖𝑓 𝑥 >
134(𝑞2−112.4)+112.4(134−𝑞1
)<𝑥<134
(𝑞2−112.4)+(134−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 134.9
𝑆𝑒𝑣𝑒𝑟𝑒(𝑥)
=
{
0, 𝑥 ≤ 128.2
𝑥 − 128.2
𝑞1−128 .2
, 𝑖𝑓 128.2 ≤ 𝑥 ≤ 𝑞1
1, 𝑖𝑓 𝑞1 ≤ 𝑥 ≤ 𝑞2
148.5 − 𝑥
148.5 − 𝑞2
, 𝑖𝑓 𝑞2 ≤ 𝑥 ≤ 148.5
0, 𝑖𝑓 𝑥 > 148.5
𝜇𝐷𝑖𝑎𝑠𝑡𝑜𝑙𝑖𝑐𝐵𝑃(𝑆𝑒𝑣𝑒𝑟𝑒)(𝑥)
=
{
0, 𝑥 ≤ 131
𝑥 − 131
𝑞2−131
, 𝑖𝑓131 ≤ 𝑥 ≤
146(𝑞2−131)+131(146−𝑞1
)
(𝑞2−131)+(146−𝑞1
)
148.5 − 𝑥
148.5 − 𝑞2
, 𝑖𝑓 𝑥 >
146(𝑞2−131)+131(146−𝑞1
)<𝑥<146
(𝑞2−131)+(146−𝑞1
)
0, 𝑖𝑓 𝑥 ≥ 148.5

55
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
Table 3. Membership function values for Shock Prediction.
In this paper, 900 fuzzy rules are defined based on (4) and part of the rules are presented in Fig. 5
while sample rule is given as;
If HeartRate is low and Respiratory Rate is normal and Temp is very high and Blood Pressure is high
THEN Prediction is high.
Fig. 5. Rule base for cardiac patient shock level prediction.
Given the crisp input vector v = [55, 25, 7, 135], their degree of memberships is calculated from the
respective triangular membership functions as shown in Table 3. Table 4 presents the firing rules based
on the set input values. Rule evaluation of the firing rules 20, 21, 165, 51, 100, 101, 35 and 36 against
the fuzzy set in Table 3 yields the result in Table 5. The Karnik Mendel Type reduction model is applied
by selecting the leftmost (L) and rightmost (R) switch points. In our study, the switch point is selected
at L=5 and R = 2 and then we compute for the values of 𝑦𝑙 and 𝑦𝑟 using Eqs. (24) and (25), respectively.
The defuzzification is carried out using Eq. (26).
Linguistic Term Upper Membership Function Lower Membership Function
Low 0,0.1733,0.3 0.0282,0.173,0.271
Normal 0.2,0.3532,0.5 0.234,0.3558,0.464
High 0.4,0.5833,0.8 0.43,0.584,0.7666
Very High 0.6,0.806,1 0.632,0.806,0.9718

56
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
Table 3. Fuzzified value (fuzzy set).
Table 4. The firing rules.
The defuzzification is carried out using
Lingusitic Variable
HeartRate
[μ
1
, μ1
]
(55)
Respiratory
Rate[μ
2
, μ2
]
(25)
Body Temp.
[μ
3
,μ3
]
(7)
Systolic Blood
Pressure
[μ
4
,μ4
]
(135)
Diastolic Blood
Pressure
[μ
4
,μ4
]
(102)
UL[1.644,0.053] UN[2.067,0.197] UL[0.649,0.569] UH[0.360,1.579] UH[0.950,0.9065]
UN[0.154,0.062] UH[0.129,0.170] UVH[0.659,0.455]
Rule
No.
Firing Interval Consequent
R20
f
1
, f 1
, ] = [1.644*2.067*0.649*0.360,
0.053*0.197*0.569*1.579] = [0.360, 0.053]
[y
1
, y 1
] = L[0.149, 1.525]
R21
f
2
, f 2
, ] = [1.644*2.067*0.649*0.659,
0.053*0.197*0.589*0.455] = [0.649,0.053]
[y
2
, y 2
] =M[0.3497, 0.6598]
R165
[f
3
, f 3
, ]= [0.129*2.067*0.649*0.360,
0.170*0.197*0.569*1.579] = [0.129,0.170]
[y
3
, y 3
]= L[0.105, 0.631]
R51
[f
4
, f 4
, ]=[0.1544*2.067*0.649*0.659,
0.062*0.197*0.569*0.455]=[0.154,0.062]
[y
4
, y 4
]=L[0.15, 0.1525]
R100
[f
5
, f 5
]=[1.644*0.129*0.649*0.360,
0.053*0.170*0.569*1.579]=[0.129,0.053]
[y
5
, y 5
]=H[0.60, 0.598]
R101
[f
6
, f 6
] =[1.644*0.129*0.649*0.659’
0.053*0.170*0.569*0.455]=[0.129,0.053]
[y
6
, y 6
]= VH[0.80, 0.798]
R35
[f
7
, f 7
, ]=[2.067*0.129*0.649*0.360,
0.197*0.170*0.569*1.579]=[0.129,0.170]
[y
7
, y 7
]=L[0.148,0.1525]
R36
[f
8
, f 8
, ]=[2.067*0.129*0.649*0.659,
0.197*0.170*0.569*0.455]=[0.129,0.170]
[y
8
, y 8
]=H[0.5302, 0.5975]
yl = L∈[1,N−1]
min
∑ f̅n
y n
+
L
n=1 ∑ fn
y n
N
n=L+1
∑ f̅n + ∑ fn
N
n=L+1
L
n=1
. (26)
yr = R∈[1,N−1]
max
∑ fn
y
̅ n
+
R
n=1 ∑ f̅n
y
̅ n
N
n=R+1
∑ fn + ∑ f̅n
N
n=R+1
R
n=1
. (27)

57
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
The
5| Results and Discussion
In this paper, experimental results for prediction problem as applied to predicting shock level in
cardiac patients using interval type-2 fuzzy logic system have been presented. The effectiveness and
generalization capability of IT2FLS have been tested using 1000 datasets obtained from University of
Uyo Teaching Hospital and Federal Medical Centre, Yenagoa all in Nigeria. Table 1 shows part of the
details of the cardiac patient’s health datasets. Five cardiac health variables namely: systolic blood
pressure, diastolic blood pressure, temperature, heart rate and respiratory rate served as the input while
cardiac shock level served as the desired output. Triangular membership function was employed for
fuzzification of the input. Fuzzy inference was derived using Mamdani inference process. Investigation
of the performance of IT2FLS and that of its type-1 counterpart were carried out. Statistical evaluation
was carried out using performance metrics of Mean Absolute Difference (MAD), Mean Absolute
Percentage Error (MAPE), Mean Square Error (MSE) and Root Mean Squared Error (RMSE),
respectively. Fig. 6 shows the plots of transformed cardiac patients’ health dataset for (a) blood
pressure diastolic, (b) blood pressure systolic, (c) temperature, (d) respiratory rate, and (e) heart rate,
respectively.
(a)
yl =
f̅1
y 1
+ f̅2
y 2
+ f̅3
y 3
+ f̅4
y 4
+ f̅5
y 5
+ f 6
y 6
+ f 7
y 7
+ f 8
y 8
f̅1 + f̅2 + f̅3 + f̅4 + f̅5 + f 6 + f 7 + f 8
. (28)
yr =
f 1
y
1
+ f 2
y
2
+ f̅3
y
3
+ f̅4
y
4
+ f̅5
y
5
+ f̅6
y
6
+ f̅7
y
7
+ f̅8
y
8
f 1 + f 2 + f̅3 + f̅4 + f̅5 + f̅6 + f̅7 + f̅8
. (29)
yl =
0.360 ∗ 1.525 + 0.649 ∗ 0.6598 + 0.129 ∗ 0.631 + 0.154 ∗ 0.1525
+ 0.129 ∗ 0.598 + 0.053 ∗ 0.798 + 0.170 ∗ 0.1525 + 0.170 ∗ 0.5975
0.360 + 0.649 + 0.129 + 0.154 + 0.129 + 0.053 + 0.170 + 0.170
= 1.9997.
yr =
0.053 ∗ 0.149 + 0.053 ∗ 0.3497 + 0.129 ∗ 0.105 + 0.154 ∗ 0.15
+ 0.129 ∗ 0.60 + 0.129 ∗ 0.80 + 0.129 ∗ 0.148 + 0.129 ∗ 0.5302
0.053 + 0.053 + 0.129 + 0.154 + 0.129 + 0.129 + 0.129 + 0.129
= 0.36593.

58
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
(b)
(c)

59
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
(d)
(e)
Fig. 6. Plots of the transformed cardiac patients’ health dataset for (a) blood pressure diastolic, (b) blood
pressure systolic, (c) temperature, (d) respiratory rate, (e) heart rate.

60
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
5.1| Interval Type-2 Fuzzy Logic Results for Remote Vital Signs Monitoring and
Shocks Level Prediction
With the fuzzy rule-base of Fig. 4 and both the input and output membership functions plots are
depicted in Fig. 7(a-e), respectively. Applying the IT2FLC to the proposed problem, we obtain the
system’s response. Fig. 8 gives the prediction results for IT2FLS. Fig. 9 presents the system’s response
of the IT2FLS for the actual and predicted shock level.
(a)
(b)

61
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
(c)
(d)

62
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
(e)
d)
)
Fig. 7. Inputs and output membership functions plot for (a) bp-diastolic (b) bp-systolic (c)
temperature (d) respiratory rate (e) heart rate and (f) shock level prediction.

63
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
Fig. 8. The prediction results for (a) IT2FLS.
Fig. 9. System’s response of the IT2FLS for the actual and predicted shock level.
5.2| Type-1 Fuzzy Logic Results for Remote Vital Signs Monitoring and Shocks
Level Prediction
Fig. 10 gives the prediction results for T1FLS while Fig. 11 presents the system’s response of the
T1FLS for the actual and predicted shock level. Fig. 12 shows the plots of the system’s response for
the actual shock level and predicted shock level for T2FLS and T1FLS.

64
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
Fig. 10. The prediction results for T1FLS.
Fig. 11. System’s response of T1FLS for the actual and predicted shock level.
5.3| Performance Evaluation
Fig. 12 and Table 5 give the results of the performance of IT2FLS with that of T1FLS where, shock
level-actual is the actual shock level, Shock Level-Predicted-IT2FLS is the predicted shock level by
IT2FLS and Shock Level-Predicted-T1FLS is the predicted shock level by T1FLS, respectively.
Statistical analysis for comparison is also made between IFLS and IT2FLS. Two experiments are
conducted in order to explore these analyses. In each case, the performance metrics are the MSE
and the RMSE. Figs. 13-15 show the plots of the two models using their test MSEs and RMSEs. Figs.
13 (a) and (b) give the prediction errors for IT2FLS and T1FLS, while the prediction accuracy of
IT2FLS and T1FLS are shown in Figs.14 (a) and (b). Figs. 15 (a) and (b) gives the error models against
error values for IT2FLS and T1FLS, respectively. Table 6 compares the performance of IT2FLS with
that of T1FLS with respect to MSE and RMSE. The results of the statistical comparison of the model
performance are presented in Table 6.

65
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
Fig. 12. Plots of the system’s response of actual shock level and predicted shock level for IT2FLS and
T1FLS.
From Fig. 12 it is observed that with the use of IT2FLS, the model is able to model uncertainty
adequately in predicting shock level of a cardiac patient and in many applications better than the
T1FLS. Also, with MFs that are intervals, the IT2FLS is able to model uncertainty in predicting shock
level of a cardiac patient in many applications better than T1FLS which MFs are not represented as
intervals values.
6| Conclusion
This paper investigated the predictive capability of Interval Type-2 Fuzzy Logic System (IT2FLS)
based on Mamdani fuzzy inference and applied to in monitoring and predicting shock level in cardiac
patients. Implementation of conventional Type-1 Fuzzy Logic System (T1FLS) was carried out for
the purpose of comparison. By the use of IT2FLS, we have been able to predict different shock levels
for cardiac patients. Specifically, the following conclusions are made:
 The IT2FLS copes with more information and handle more uncertainties in health data.
 The IT2FLS performs significantly well compared to TIFLS with model errors of 0.0005 against 0.079.
 The IT2FLSs with interval MF can reduce the effects of uncertainties in most health applications.
In the future, we intend to explore Tagaki Sugeno Kang (TSK) fuzzy inference to conduct more
experiments using the same data sets. Also, we intend to optimize our system using flower pollination
algorithm for performance improvement. More so, we will apply other fuzzy modeling functions such
as triangular and trapezoidal functions, respectively.

66
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
Table 5. Results of shock level actual and predicted for T2FLS and T1FLS.
6.1| Ethical Issues
Ethical issues came to play in this research as the research involved gathering data from
cardiovascular patients’ records. However, the research was not involved in the direct collection of
the data, but a review of patients’ files and medical histories after due permission was granted by the
responsible authorities. Hence, we discuss the ethical issues under two areas:
S/No. Shock Level-
Actual
Shock Level-Predicted
IT2FLS
Shock Level-Predicted
T1FLS
1. 0.430209738 0.430534538 0.443508938
2. 0.661076429 0.661401229 0.674375629
3. 0.600932789 0.601582389 0.614556789
4. 0.382681769 0.383006569 0.391656169
5. 0.492674633 0.492999433 0.501649033
6. 0.284818273 0.285467873 0.289792673
7. 0.381977912 0.382627512 0.408576312
8. 0.578692885 0.579342485 0.583667285
9. 0.463544732 0.463869532 0.481168732
10. 0.456734044 0.457383644 0.483332444
11. 0.604593743 0.605243343 0.613892943
12. 0.464519101 0.464843901 0.473493501
13. 0.192299351 0.192948951 0.210248151
14. 0.369657904 0.370307504 0.374632304
15. 0.28822936 0.28887896 0.30185336
16. 0.650264118 0.650588918 0.672212918
17. 0.734445205 0.734770005 0.760718805
18. 0.186542266 0.187191866 0.204491066
19. 0.564212161 0.564861761 0.573511361
20. 0.518631549 0.518956349 0.544905149
21. 0.408237197 0.408561997 0.421536397
22. 0.659320718 0.659645518 0.672619918
23. 0.400810079 0.401134879 0.422758879
24. 0.403371577 0.404021177 0.408345977
25. 0.410957191 0.411606791 0.437555591
26. 0.408079746 0.408729346 0.426028546
27. 0.332620921 0.333270521 0.359219321
28. 0.350599807 0.351249407 0.368548607
29. 0.46014474 0.46079434 0.46944394
30. 0.296303978 0.296628778 0.309603178
31. 0.492720862 0.493045662 0.506020062
32. 0.40598575 0.40631055 0.43225935

67
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction
 Consent form. The consent form through written permission was obtained from the health authority before embarking
on the research. A sample of the authorization clearance is used by the ethical committee of Federal Medical Centre,
Yenagoa and University Teaching Hospital, Uyo
 Data protection. Data protection is ensured by not revealing patients’ personal details such as Name, Address,
Occupation and many others. Hence, data gathered excluded this information.
Table 6. Comparison of T1FLS and IT2FLS in shock level prediction.
References
Zadeh, L. A. (1965). Fuzzy sets. Inf. Control, 8, 338–353.
Negnevitsky, M. (2005). Artificial intelligence: a guide to intelligent systems. Pearson education.
Chiu, S. (1997). Extracting fuzzy rules from data for function approximation and pattern classification.
In D. Dubois, H. Prade, and R. Yager (Eds.) Fuzzy information engineering: a guided tour of applications. John
Wiley & Sons.
Mamdani, E. H. (1974, December). Application of fuzzy algorithms for control of simple dynamic plant.
In Proceedings of the institution of electrical engineers (Vol. 121, No. 12, pp. 1585-1588). IET.
Ramkumar, V., Mihovska, A. D., Prasad, N. R., & Prasad, R. (2010). Fuzzy-logic based call admission
control for a heterogeneous radio environment. The 12th international symposium on wireless personal
multimedia communications (WPMC 2009). Japan, Sendai, Japan.
https://vbn.aau.dk/en/publications/fuzzy-logic-based-call-admission-control-for-a-heterogeneous-radi
Selvi, M. P., & Sendhilnathan, S. (2016). Fuzzy based mobility management in 4G wireless
Networks. Brazilian archives of biology and technology, 59(SPE2).
Mali. G. U. (2017). Fuzzy based vertical handoff decision controller for future networks. International
journal of advanced engineering, management and science (IJAEMS), 3(1), 111-119.
Abbasi, R., Bidgoli, A., & Abbasi, M. (2012). A new fuzzy algorithm for improving quality of service in
real time wireless sensor networks. International journal of advanced smart sensor network systems
(IJASSN), 2(2), 1-14.
Dogman, A., Saatchi, R., & Al-Khayatt, S. (2012). Quality of service evaluation using a combination of
fuzzy C-means and regression model. World academy of science, engineering and technology, 6, 562-571.
Umoh, U. A., & Inyang, U. G. (2015). A fuzzfuzzy-neural intelligent trading model for stock price
prediction. International journal of computer science issues (IJCSI), 12(3), 36.
Umoh, U., & Asuquo, D. (2017). Fuzzy logic-based quality of service evaluation for multimedia
transmission over wireless ad hoc networks. International journal of computational intelligence and
applications, 16(04), 1750023.
Mendel, J. M., & John, R. B. (2002). Type-2 fuzzy sets made simple. IEEE transactions on fuzzy
systems, 10(2), 117-127.
Zadeh, L. A. (1975). The concept of a linguistic variable and its application to approximate reasoning—
I. Information sciences, 8(3), 199-249.
Hagras, H. (2007). Type-2 FLCs: A new generation of fuzzy controllers. IEEE computational intelligence
magazine, 2(1), 30-43.
Tan, W. W., & Chua, T. W. (2007). [Review of Uncertain rule-based fuzzy logic systems: introduction and
new directions, by J. M. Mendel]. IEEE Computational intelligence magazine, 2(1), 72-73.
Postlethwaite, B. E. (1993). [Review of Fuzzy control and fuzzy systems, by W. Pedrycz]. International
journal of adaptive control and signal processing, 5(1), 87-153. https://doi.org/10.1002/acs.4480050208
Models Performance MSE RMSE
T1FLS Prediction Error 0.004 0.018
Model Error 0.0003 0.079
Prediction Accuracy 0.9997 0.9821
IT2FLS Prediction Error 0.0001 0.0006
Model Error 2.57E-07 0.0005
Prediction Accuracy 1 0.9995

68
Umoh
et
al.
|J.
Fuzzy.
Ext.
Appl.
2(1)
(2021)
41-68
Hagras, H. A. (2004). A hierarchical type-2 fuzzy logic control architecture for autonomous mobile
robots. IEEE transactions on fuzzy systems, 12(4), 524-539.
Mendel, J. M., John, R. I., & Liu, F. (2006). Interval type-2 fuzzy logic systems made simple. IEEE
transactions on fuzzy systems, 14(6), 808-821.
Zimmermann, H. J. (2011). Fuzzy set theory—and its applications. Springer Science & Business Media.
Castillo, O., & Melin, P. (2012). Recent advances in interval type-2 fuzzy systems. Springer.
Isizoh, A. N., Okide, S. O., Anazia, A. E., & Ogu, C. D. (2012). Temperature control system using
fuzzy logic technique. International journal of advanced research in artificial intelligence, 1(3), 27-31.
Roy, N., & Mitra, P. (2016). Electric load forecasting: an interval type-ii fuzzy inference system based
approach. International journal of computer science and information technologies (IJCSIT), 7(6), 2515-
2522.
Umoh, U. A., Inyang, U. G., & Nyoho, E. E. (2019). Interval type-2 fuzzy logic for fire outbreak
detection. International journal on soft computing, artificial intelligence and applications (IJSCAI), 8(3), 1-
20.
Castillo, O., Melin, P., Kacprzyk, J., & Pedrycz, W. (2007, November). Type-2 fuzzy logic: theory and
applications. 2007 IEEE international conference on granular computing (GRC 2007) (pp. 145-145). IEEE.
Mamdani, E. H., & Assilian, S. (1975). An experiment in linguistic synthesis with a fuzzy logic
controller. International journal of man-machine studies, 7(1), 1-13.
Mamdani, E. H. (1976). Advances in the linguistic synthesis of fuzzy controllers. International journal
of man-machine studies, 8(6), 669-678.
Umoh, U. A., Nwachukwu, E. O., Obot, O. U., & Umoh, A. A. (2011). Fuzzy-neural network model
for effective control of profitability in a paper recycling plant. American journal of scientific and
industrial research (AJSIR), 2(4), 552-558. DOI: 10.5251/ajsir.2011.2.4.552.558
Wu, H., & Mendel, J. M. (2002). Uncertainty bounds and their use in the design of interval type-2
fuzzy logic systems. IEEE Transactions on fuzzy systems, 10(5), 622-639.
Wu, D. (2005). Design and analysis of type-2 fuzzy logic systems (Master’s Thesis, Department of
Electrical and Computer Engineering, National University of Singapore). Retrieved from
https://core.ac.uk/download/pdf/48633921.pdf
Wu, D., & Tan, W. W. (2005, May). Computationally efficient type-reduction strategies for a type-2
fuzzy logic controller. The 14th IEEE international conference on fuzzy systems, 2005. FUZZ'05. (pp.
353-358). IEEE.
Karnik, N. N., & Mendel, J. M. (2001). Centroid of a type-2 fuzzy set. Information SCiences, 132(1-4),
195-220.

69
Interval
type-2
fuzzy
logic
system
for
remote
vital
signs
monitoring
and
shock
level
prediction

Interval type-2 fuzzy logic system for remote vital signs monitoring and shock level prediction

More Related Content

What's hot (17)

Similar to Interval type-2 fuzzy logic system for remote vital signs monitoring and shock level prediction (20)

More from Journal of Fuzzy Extension and Applications (20)

Recently uploaded (20)

Interval type-2 fuzzy logic system for remote vital signs monitoring and shock level prediction