Annotated Bibliography Assignment PDF

Link of blog/website: http://thinkspace.csu.edu.au/asjadprojectblog/
Full Name MOHAMED ASJAD HILMY MOHAMED HILMY ISMAIL
Student ID 11614177
Subject ITC571 – Emerging Technology and Innovations
Assignment No Annotated Bibliography
Due Date 26/04/2018
Lecturer’s Name Malka N. Halgamuge
Review paper for, “How different types of
devices available to detect Epilepsy:
Seizure.”
Mohamed Asjad Hilmy Mohamed Hilmy Ismail
[e-mail]
School of Computing and Mathematics, Charles Sturt University, Melbourne, Victoria
I. ABSTRACT
Epilepsy is a foam of medical condition which a person has recurrent seizures. In this study
will be discussed how the technology helps for all the patients who gets seizures. A seizure is
known as an abnormal, discharging the nerve cells of the brain in an informal order [1]. About
33% of patients with epilepsy keep on having seizures regardless of ideal medical solution.
Systems utilized to distinguish seizures may can possibly enhance results in these patients by
permitting more custom fitted treatments and might, furthermore, have a part in mischance and
SUDEP prevention. Automated seizure recognition and prediction require computation which
utilize include calculation and consequent order. In the course of the most recent couple of
decades, techniques have been created to distinguish seizures using scalp and intracranial EEG,
electrocardiography, accelerometer and movement sensors, electro-dermal action, and
sound/video captures. To date, it is misty which blend of identification advances yields the best
1

outcomes, and methodologies may at last should be individualized. This review introduces a
diagram of seizure discovery and related expectation techniques and talks about their potential
uses in closed-loop warning systems in epilepsy [2].
II. INTRODUCTION
Epilepsy is a medical condition that influences a person’s mind activity. This can prompt
seizures and lead to serious conditions. Epilepsy is regularly analyzed in patients, making it
difficult for guardians/parents to screen their patients’ seizures constantly. Epilepsy cannot be
cured by the technology and should be followed by appropriate medication. Since epilepsy
attacks are undetectable, there are detecting devices have been developed such as EEG and
Electrocorticography, Baby monitors, ECG, Accelerometry, Video detecting systems, Mattress
sensors and so on to predict and detect and then alert to appropriate person including the patients
[3] [2]. Apparently seizure attacks cannot be predicted, all the patients must be followed by
appropriate medication but by having developed devices to predict the epilepsy attacks is good
because at least it could be prevented and be alerted. The need of this topic is to exhibit to public
that how should it be helpful for those who suffer from this kind of disease and how the
developed technology to be utilized.
Until this century, pharmacological solutions and brain surgery were the main two choices
accessible to patients with probably the most widely recognized neurological issue, for an
instance, epilepsy and Alzheimer’s illness. Medications regularly have restricted general
treatment viability, usually cause symptoms, and the patient creates protection from them
inevitably. Furthermore, just a few patients are regarded to be hopefuls for mind surgery
(example 15% for epilepsy). This is because for numerous cases, the wellspring of neurological
issue is near an essentially imperative part of the brain where a surgery could come about in
irreversible harm to patient’s fundamental functionalities of body. Since after Numerous
research groups have been dealing with both enhancing and calibrating the innovation for those
diseases, also creating and validating an extensive therapeutic device for conclusion and
treatment of different other neurological diseases too [4].
2

A worry for a patient with epilepsy isn’t just the seizures that are seen, however those that go
undetected. This is particularly valid for seizures a patient may have in their sleep. The objective
of epilepsy treatment is to utilize medicines and different treatments to keep a patient seizure
free. Nonetheless, it’s conceivable a patient could think their epilepsy is controlled, however
have seizures around evening time. Another worry about seizures is the danger of sudden
unforeseen death in epilepsy (SUDEP). This happens when a man passes away all of a sudden
after a seizure. Despite the fact that the correct causes are obscure, changes in breathing, (for an
instance, something choking out the individual) or heart rhythms can be a factor. By recognizing
seizures, devices for epilepsy might have the capacity to prevent SUDEP [5].
By figuring out why this paper would be significant, because this paper will bring all the
technology/devices [2] available in the market now and how to overcome SUDEP. At the same
time there might be side effects in these devices too. Finally aim is to propose to public that what
are the devices available to detect epilepsy and what are the new features can be added to
existing developments.
In addition to this, it can be said that present research is looking forward to have
identification of diverse norms. In previous studies the experts has not covered all the factors and
information related to Epileptic. Nonetheless, it has been seen in past investigations that epilepsy
is a kind of the neurological problem, which can cause brokenness in the mind because of
disgraceful working of nerve cell. It comes about into rehashed seizures, which prompts unusual
development of body parts, for example, hands, legs and head. Albeit, key explanation for the
recognizable proof of Epileptic. Additionally, the techniques utilized as a part of the
investigations are not suitable which has likewise affected the working in differing way. Epilepsy
is neurological disarranges caused by unending brokenness of the cerebrum and its
powerlessness to produce unjustifiable and erratic seizures. But the ways to deal with such issues
were not explained in the studies. It has created conflict in understanding of the concept. By
having an evaluation of such values the gape in previous studies can be overcome. It has also
been identified that these individuals are normally confined amid the night and defenseless
against a few physical wounds or asphyxia because of a blocked aviation route subsequent to
gulping their tongues.
Present study key aim is to have development of learning in respect to Epileptic Devices.
By having an understanding about the subject the issues in terms of health can be resolved in
3

desired manner. Along with this, it has been noticed that factors and symptoms regarding
Epileptic can easily be evaluated in desired manner. With an effective consideration of such
measures the overall practice can be advanced critical manner. Another key motive of study is to
make understand that the objective of epilepsy treatment is to utilize medicines and different
treatments to keep a patient seizure free. It helps in understanding the key values related to
Epileptic
III. MATERIAL AND METHODS
A. Data Collection and Data Inclusion Criteria
Events of various data collection methods’, data was gathered from eleven research
papers distributed by past researchers in the field of medical science and filtered the important
information that can be helpful for full filing this paper’ output. In order to have effective data
collection about the subject the experts has focused on secondary data collection. In this, critical
annotated bibliography has been presented to make sure about the study data collection.
Below table has been drawn to assess the data inclusion criteria, and the attributes are
Author, Technology/Device, Methods, Category/Classification and Other Considerations. Along
with this, consideration of diverse methods used in different studies has also been evaluated in
order to make sure that appropriate information has been collected. Methods like open loop and
random collection are major concerned of experts which has allowed to attain better information
about the subject. Data collected is also being presented in order to make sure that which kind of
methods can be employed to deal with the neurological issue.
Author Technology/Device Methods Category/Classification Other Consideration
M.Tariqus,
Salam;
Tonekaboni,
Sana ; Kassiri,
Hossein [4]
Neurostimulators Open-loop &
Closed-loop
Neurostimulators
Implanted Advisory
System
 Higher
computation
coat
 User friendly
environment
of
programming
 Surgical
could damage
the patients’
4

body
functionality
Shao, Weiwei
; Miao,
Yuanmin ; Li,
Zhangjian [6]
Implantable
ultrasound device
Transducer and
Phantom Design
Implanted Advisory
System
 Surgical
could damage
the patients’
body
functionality
 Uses 3D
technology
Abdelhalim,
Karim ;
Jafari, Hamed
Mazhab;
Kokarovtseva,
Larysa [7]
Neural Synchrony
monitoring
wireless
technology
1. System-on-
chip VLSI
Architecture
2. Offline
human early
Seizure
detection
3. Online in
VIVO
Rodent
Seizure
Control
Implanted Advisory
System
 Surgical
could damage
the patients’
body
functionality
Manzouri,
Farrokh ;
Schulze-
Bonhage,
Andreas ;
Dümpelmann,
Matthias [8]
Optimized
Detector for
implanted device
Random Forest
Classifier
Implanted Advisory
System
 Surgical
could damage
the patients’
body
functionality
 Used random
forest
classifier to
increase the
5

performance
 High
sensitivity
V. Tonpe,
Snehal ; G.
Adhav,
Yashwant ; K.
Joshi, Atul [9]
Seizure detection
using Micro Sensor
3-axis
accelerometer
(MEMS based)
Accelerometry  Minimal
computational
energy
 Low cost
 High
reliability
Salem, Osman
; Rebhi,
Yacine ;
Boumaza,
Abdelkrim
[10]
Wireless 3-D
Accelerometer
Sensors
Not mentioned Accelerometry  Computation
complexity
Wu, Ge ; Xue,
Shuwan [11]
Portable Pre-
impact Fall
Detector
The inertial frame
velocity profile of
the body
Accelerometry  Threshold is
successful of
decreasing
the false
alarm
Carlson, Chad
; Arnedo,
Vanessa ;
Cahill,
Maria ;
Devinsky,
Orrin [12]
MP5 Monitor Not mentioned Mattress Sensors  Negative
alarms <
positive
alarms
Conradsen,
Isa ;
Beniczky,
Wearable sEMG Not mentioned Accelerometry  High
sensitivity
with low false
6

Sándor ;
Wolf, Peter
[13]
alarms
Adwitiya,
Aziis Yudha;
Hareva, David
Habsara;
Lazarusli,
Irene Astuti
[14]
Epileptic Alert
using Smart Phone
Motion Sensors Not Declared  Smart phones
use motion
senses
Lin, Shih-Kai;
Lin, Yu-Shan;
Lin, Chin-
Yew [15]
Smart Headband Smart Device
APP’s integrated
with cloud
computer platform
Accelerometry  Low
computational
cost
B. Analysis
Epilepsy is a common neural issue ailment. Most patients use antiepileptic meds to
decrease their seizures, however just about 33% of the patients are calm safe epilepsy. The
elective treatment is the resection surgery of emptying the epileptogenic zone. Nevertheless,
every above patient will at introduce have a couple of seizures in the midst of their step by step
life, which will affect the patients' close to home fulfillment, and further familiarize dangers and
weight with patients and people around. For a productive collaboration, related cerebrum
activities and events should be reliably perceived using distinctive strategies including machine
learning frameworks. To this end, a combination of trademark hail incorporates and furthermore
one of a kind sorts of classifiers can be used. One possible use of such an association is for
epilepsy patients. A novel approach for the get-together of patients with difficult to treat epilepsy
is the usage of electrical prompting first and foremost times of the seizure age in a close circle
way which can be recognized in an insert.
Wearable sensors enable whole deal incessant physiological watching, which is basic for
the treatment and organization of various unending sicknesses, neurological disseminates, and
mental health issues. A psychological lopsidedness Spectrum Disorders (ASD) – there is
growing eagerness for early-age ID of ASD and to improve medicines. Regardless of the way
7

that EDA is regularly assessed from the fingertips, imperative EDA data can in like manner be
accumulated from the wrists and lower legs, which are the favored regions for versatile
estimation of EDA. Epilepsy is the third most normally dissected neurological issue, which is
depicted by reoccurring seizures.
IV. Results
The data has been focusing on the feature of the scaling results with proper evaluation of
the feature scaling that is based on evaluating the performance for the random forest classifier.
Here, the focus is on the classifiers with the selected channels for the patients and then averages
which are set over for all the patients. The detection of the time window is set for the
overlapping of the two seconds with the one second.
On Average Performance
Measures Feature Scaling No Features
Sensitivity Analysis 0.93 0.89
Delayed detection 13.15 11.33
FDR 6.33 5.67
The selection is based on how the patient from one channel with the minimized delayed
detection is compared with the classifier that holds the average performances for the different
patients. The selections are based on the shorter delays of detection with closures that are related
to work on the onset zones with providing the results. The selection is for the channels that tend
to provide the relevant information with proper evaluation of the classifier performance.
Measures Feature Scaling No Features
Sensitivity Analysis 0.95 0.94
Delayed detection 7.87 5.42
FDR 5.78 5.64
For the single channel performance
With the detection time window optimization results, there are different seizer detection results
that are using one second detection time window which is important for the non-overlapping
with using the two second detection that is set for the time windows. The average is determined
8

with the classifier with the selected channels for the different patients and then average for all the
patients.
Measures Feature Scaling No Features
Sensitivity Analysis 0.88 0.89
Delayed detection 12.65 11.33
FDR 6.62 5.67
The detection for the window time with the single channel performance is based on the detection
and the time activities.
Here, to evaluate the effects of the results, there is a need of the normalization of the data which
is then compared to the different results for the seizure detection. This is when there is a need for
the normalization which is based on the median decaying memory and then handling the
statistical normalization methods as well. The results are detected through without any data
normalization where there is a need to set the detection window patterns with the time of the one
second overlapping [8]. The comparison is done to the average performances and then handling
the time which is about how the classifier is able to select the channels for the different patients
and then average it for all the other patients as well.
The below image displays the normalization effect on the average performance where:
Measures No Normalization Median Decaying Statistical
Sensitivity Analysis 0.89 0.91 0.88
Delayed detection 11.33 14.03 12.67
FDR 5.67 5.94 5.4
The standards are set for analyzing the detection delays from the different patients who are then
compared to the classier with the average performance for the patients.
The results are based on the different electrode montages with the seizure detection performance
that works with the selected channels. The forms are set with the average results where the
classifier is depending upon how the patients and the average over all the patients is set. The time
window is set depending upon the one second overlapping [8]. Considering the developed
system, there are implantation process which includes the expectations for the use of the
9

wearable accelerometer where the sensor detection is for the seizure that is found to be effective.
The effectiveness is based on how the devices are subjected to handle the unattended patient,
with impact that occurs at night which is found to be unobserved. This can easily be avoided
through the use of the seizure detection system with the triggering of the alarms that one is aware
of the care taken with proper actions.
Considering the robust seizure detection algorithm, it has been seen that there are long term
recording which are mainly for preserving the performances. It includes the handling of the
delaying in the preservation of high sensitive with the keeping of different detection which is
lower and important as well. The assessment is based on the approaches which are based on the
evaluation of the techniques that are worked upon through handling the seizure detection
approaches. The changing properties are for the bio signals over the time with the idea related to
the form that include the feature scaling methods [8]. The classifications include the figure of
how one can work with the FDR which tends to increase and then handled through the seizure
detection method. The FDR increases slightly with the results that confirm to the features that are
based on performance measures. The optimization approach is based on searching and working
on the optimum detection time window which highlights about showing the time windows for
the two seconds and then overlapping the overall performance.
One need to work on the different frame velocity profiles which includes the profitability
standards that include the impact of working on the optimization approach which includes the
detection with the time windows for the seizure detection. Here, the results are defined for the
time window of the two second with one second that is overlapping with the better performance.
The forms are related to how the analysis is worked upon with the sensitive with two second of
the time window that is overlapping with the detection delays that are about 1.3 second and the
FDR is also seen to be smaller comparatively.
V. DISCUSSION
The principle point of this research is to distinguish different devices/components are
being used as a part of the field of medical and investigate those through a beneficial procedure
with a specific end goal to serve significant result from the research which would be very
effective for future to improve the existing developments. There are several devices have been
developed to prevent epilepsy though only few of them being used by the industry and running
successfully. All the identified technologies can be classified in to several categories as
10

mentioned in the above table. It has been recognized that roughly 50 million individuals
worldwide are epileptic. Epilepsy is the third most regular neurological issue following stroke
and Alzheimer's illness, however it forces higher expenses on society than stroke does. Epileptic
seizure onsets are regularly described by particular forerunners incorporating expanded variances
in stage synchrony and resulting expanded cerebrum synchronization. The early seizure location
calculation depends on figuring the greatness, stage and stage synchrony of neural flags in
particular recurrence groups in the neural flag range. At the point when the processed stage
synchrony increments over a programmable edge in a moving normal time window, a seizure is
distinguished. There are three kinds of ready gadgets accessible today in light of movement
discovery for epilepsy, for example, sleeping cushion gadgets, watch gadgets, and camera
gadgets. Bedding gadgets are normally set under a sleeping pad to recognize vibrations.
What's more, one might say that Epileptic caution applications made utilizing the
Android Software Development Kit (Android SDK) with focused android Smartphone gadgets
outfitted with GPS, movement sensor and position sensor. Most complex cell phones have
worked in sensors that can be utilized to gauge movement, introduction, and changing ecological
conditions. By having an effective consideration of diverse technological tools the issues in
terms of epileptic can be overcome. Pharmaceuticals routinely have confined general conduct
practicality, as a rule cause side effects, and the patient makes security from them unavoidably.
Besides, only a couple of patients are respected to be hopefuls for mind surgery. A stress for a
patient with epilepsy isn't only the seizures that are seen, however those that go undetected. This
is especially substantial for seizures a patient may have in their rest. The target of epilepsy
treatment is to use solutions and diverse medications to keep a patient seizure free. Regardless,
it's possible a patient could think their epilepsy is controlled, however have seizures around night
time.
VI. Conclusion
The overall work focus on the epileptic seizure detection which is based on the detection
through proper performance based on the biological signal that is coming from the human skin. It
includes the combination of the fabrics and the use of the pre-amplifier that is set into the analog
that is coming from the front end-circuit with the assembling on the flexible printing circuit.
11

They are for fitting with the textile and then handling the designing which is important for the
detection tag seizure [15]. It includes the mixed signals with the BLE chip for the wireless
transmission. The validation of the epilepsy is based on the utilization of the single implantable
electronic microchip. This is for the experimental procedures with the detection and the closed
loop seizure where the demonstrations are about handling the in vivo on rats. The computation is
based on the magnitude with the phases and the phase synchrony. The operations are based on
the neural stimulatory with the 64 channels that are depending upon how the closed loop
treatment is used for the integrable epilepsy with 80% efficiency as well.
Despite of the limitations for the studying patients, there are certain impatient setting
where the study is about focusing on the MPS devices [12]. The patients tend to believe about
the tonic clonic seizures. Here, the study is about the utilization of the calibrated devices which
are for the home setting and then for the monitoring as well. They will be able to characterize it
in a proper manner as well.
12

VII. REFERENCES
[1] J. S. Huff and M. Huff, "Epilepsy," [Online]. Available:
https://www.emedicinehealth.com/epilepsy/article_em.htm#what_is_epilepsy.
[2] S. Ramgopal and S. Thome-Souza, "Seizure detection, seizure prediction, and closed-loop warning
systems in epilepsy," Epilepsy & Behavior, vol. 37, pp. 291-307, August 2014.
[3] Elsayed, Saad Zaghloul and Bayoumi, "BCI/AIS Low Power Adaptive Architecture for Early
Prediction of epilepsy seizrues," 2017.
[4] S. M.Tariqus, S. Tonekaboni and H. Kassiri, "Closed-Loop Neurostimulators: A Survey and A Seizure-
Predicting Design Example for Intractable Epilepsy Treatment," IEEE TRANSACTIONS ON
BIOMEDICAL CIRCUITS AND SYSTEMS, vol. 11, pp. 1026 - 1040, 2017.
[5] R. Nall, "Bracelets and Devices for People with Epilepsy," 29 July 2016. [Online]. Available:
https://www.healthline.com/health/bracelets-and-devices-epilepsy.
[6] W. Shao, Y. Miao and Z. Li, "An Intracranial Implantable Ultrasound Device for Seizure Mapping," in
Ultrasonics Symposium (IUS), 2017 IEEE International, 2017.
[7] K. Abdelhalim, H. M. Jafari and L. Kokarovtseva, "Neural Synchrony-Monitoring Wireless Brain
Implant for Intractable Epilepsy Neuromodulation," San Diego, CA, 2013.
[8] F. Manzouri, A. Schulze-Bonhage and M. Dümpelmann, "Optimized Detector for Closed-loop
Devices for," in Man, and Cybernetics (SMC), 2017.
[9] S. V. Tonpe, Y. G. Adhav and A. K. Joshi, "Epileptic Seizure Detection using Micro Sensor," Chennai,
2017.
[10] O. Salem, Y. Rebhi and A. Boumaza, "Detection of Nocturnal Epileptic Seizures Using Wireless 3-D
Accelerometer Sensors," Natal, Brazil, 2014.
[11] G. Wu and S. Xue, "Portable Preimpact Fall Detector With Inertial Sensors," IEEE TRANSACTIONS
ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, vol. 16, pp. 178 - 183, 03 04 2008.
[12] C. Carlson, V. Arnedo, M. Cahill and O. Devinsky, "Detecting nocturnal convulsions: Efficacy of the
MP5 monitor," Seizure, vol. 18, pp. 225-227, April 2009.
[13] I. Conradsen, S. Beniczky and P. Wolf, "Evaluation of novel algorithm embedded in a wearable
sEMG device," San Diego, CA, 2012.
[14] A. Y. Adwitiya, D. H. Hareva and I. A. Lazarusli, "Epileptic Alert System on Smartphone," in Soft
Computing, Intelligent System and Information Technology, Denpasar, Indonesia, 2017.
[15] S.-K. Lin, Y.-S. Lin and C.-Y. Lin, "A Smart Headband for Epileptic Seizure Detection," Bethesda,
2017.
13