Eastern-European Journal of Enterprise Technologies ISSN 1729-3774 1/4 ( 121 ) 2023
56
Copyright © 2023, Authors. This is an open access article under the Creative Commons CC BY license
ANOMALY
DETECTION IN
INTERNET OF
MEDICAL THINGS
WITH ARTIFICIAL
INTILLEGENCE
Shalau Farhad Hussein
College of Information Technology
and Computer Science
Presidency of University of Kirkuk*
Zena Ez. Dallalbashi
Department of Electronic
Technical Institute
Northern Technical University
Mosul, Iraq, 41002
Ahmed Burhan Mohammed
Corresponding author
College of Dentistry Presidency
of University of Kirkuk*
E-mail: ahmedlogic79@uokirkuk.edu.iq
*University of Kirkuk
Al-Sayada Area Kirkuk, Kirkuk, Iraq, 36001
Internet of things (IoT) becomes the most popular term
in the recent advances in Healthcare devices. The health-
care data in the IoT process and structure is very sensi-
tive and critical in terms of healthy and technical consid-
erations. Outlier detection approaches are considered as
principal tool or stage of any IoT system and are main-
ly categorized in statistical and probabilistic, clustering
and classification-based outlier detection. Recently, fuzzy
logic (FL) system is used in ensemble and cascade systems
with other ML-based tools to enhance outlier detection
performance but its limitation involves the false detection
of outliers. In this paper, we propose a fuzzy logic system
that uses the anomaly score of each point using local out-
lier factor (LOF), connectivity-based outlier factor (COF)
and generalized LOF to eliminate the confusion in classi-
fying points as outliers or inliers. Regarding human acti-
vity recognition (HAR) dataset, the FL achieved a value
of 98.2 %. Compared to the performance of LOF, COF, and
GLOF individually, the accuracy increased slightly, but the
increase in precision and recall indicates an increase in
correctly classified data and that neither true nor abnormal
data is classified wrongly. The results show the increase in
precision and recall which indicates an increase in correct-
ly classified data. Thus, it can be confirmed that fuzzy logic
with input of scores achieved the desired goal in terms of
mitigating cases of false detection of anomalous data. By
comparing the proposed ensemble of fuzzy logic and diffe-
rent types of local density scores in this study, the outcomes
of fuzzy logic presents a new way of elaborating or fusing
the different tools of the same purpose to enhance detection
performance
Keywords: anomaly detection, outlier score, anomaly
score, fuzzy logic, hybrid system
UDC 621
DOI: 10.15587/1729-4061.2023.274575
How to Cite: Hussein, S. F., Dallalbashi, Z. Ez., Mohammed, A. B. (2023). Anomaly detection in internet of medical things
with artificial intillegence. Eastern-European Journal of Enterprise Technologies, 1 (4 (121)), 56–62. doi: https://doi.org/
10.15587/1729-4061.2023.274575
Received date 07.12.2022
Accepted date 15.02.2023
Published date 28.02.2023
1. Introduction
Internet of things (IoT) becomes the most popular term
in the recent advances in Healthcare devices. Whereas the
most common example of IoT in health care is remote mon-
itoring of patients, like IoT devices that collect patient data
such as heart rate and biomarkers [1]. IoT devices offer new
features for healthcare providers including accurate and in-
time patient monitoring and treating. In addition, various
wearable monitors and sensors provide plenty of benefits and
capabilities for healthcare providers and their patients. Basi-
cally, the applications of IoT in the medical field are remote
patient monitoring like biosignals, depression and mood
monitoring and human activity recognition (HAR). Reliable
IoT technology in the medical field requires proper data
transmission and addressing critical security challenges [2].
Healthcare data is very sensitive and critical and any
distortion or malfunction in the transmitter and receiver
may change or contaminate the diagnosis of the patients. In
cases of cyber-attacks or the noise of the equipment and the
surrounding environment, there will be anomalies or outliers
in this data. Those anomalies must be detected accurately in
the real-time or appropriate moment to prevent any serious
consequences, the worst of which is the life of the monitored
patient. The concept of anomaly detection includes a wide
range of applications, such as outlier or novelty detection and
intrusion detection [3].
Outlier detection approaches are considered as principal
tool or stage of any IoT system [2] and are mainly catego-
rized in statistical and probabilistic, clustering and classifi-
cation-based outlier detection. Statistical methods rely on
previously measured data to approximate the correct beha-
vior model of the data such as the autoregressive and moving
average (ARMA) prediction model, adaptive kernel density
estimator (AKDE) approach and classification and regression
Trees (CART) model where new data is compared to the pre-
viously generated model data. If the results show the statisti-
cal significance of the data is different, the new data is marked
as an anomaly. The probabilistic method relies on distinguish-
ing outliers from normal or inlier data. If the probability is
below a predetermined threshold, it is classified as abnormal.
Outlier detection and classification during monitoring and
detection of human activity using environmental sensors,
wearable sensors, or both is still under development due to
the prevalence and recent achievements in medical IoT and
6G era communications. Clustering-based outlier detection