IOT Based Anesthesia Parameters Monitoring with Doctor Decision Assistance using Machine Learning

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3509
IOT Based Anesthesia Parameters Monitoring with Doctor
Decision Assistance using Machine Learning
DR. RADHA R1, NAGA SAI RAM AELLA2
1Associate Professor, School of computer science and engineering, Vellore Institute of Technology, Chennai, India
2Undergraduate student, School of computer science and engineering, Vellore Institute of Technology, Chennai,
India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Anesthesia plays a very important role in the
surgery. In a long surgery anesthesia is given multiple times
but not at a time to avoid the high dose which may affect the
patient. If the dose is not given in time there is a chance where
the patient will be conscious and the situation like these tends
to panic. Even the anesthesia overdose may lead to deaths.
There are some factors which should be noticed to the
concerned anesthesia doctor before giving the anesthesia to
the patient. The anesthesia doctor can get assistance with the
help of advancements in Computer science by using IOT and
ML. The raspberry pi used in this project will collect the data
from the sensors and with the help of an in-built wi-fi module
in the raspberry pi the data will be transmitted to the
Thingspeak. Earlier this system was purely based on the
Arduino for just collecting the data and automatic injector
was used which is not safe for the patient as during surgery
anesthesia is given at different parts of the body based on
where the surgery is going to be done. So, It is not possible to
use the automatic injectors for anesthesia. To overcome the
dose fluctuations and to get assistance from the Machine
learning algorithm to doctor which shows the risk prediction
based on the parameters entered by the doctor. The
Immediate message will be sent to the concerned staff if there
is any fluctuation in the reading.
Key Words: IOT, ML, wi-fi, Thingspeak, Raspberry pi,
Anaesthesia assistant, Arduino, Logistic Regression.
1. INTRODUCTION
While performing long duration surgeries the anaesthesia is
given to the patient several times but it is not delivered at a
time as it may lead to overdose. Overdose may even end up
in patient deaths. As, It is given multiple times to the patient
during surgery the doctor needs to visualize the parameters
every time he needs to inject the anaesthesia.
Not only overdose but low dose than required may make
situations panic during the surgery. Here Computer Science
fields like IOT and Machine learning logistic regression
algorithm will help us to overcome the above with
Indications when the parameter values go beyondor behind
the par value and the machine learning algorithm predicts
the risk based on the parameters. It is a hassle - free for the
doctor as he gets assistance from the risk predictor. It is not
possible to monitor every value every time for the doctor
there are chances the parameter may get skipped.
The data from the sensors and the data is transmitted to the
thingspeak cloud and from the cloudthedata fromsensorsis
retrieved to the website and different visualizations are
shown based on the par values of the parameters. Even, If
the values are fluctuating below or above the threshold
immediate message will be sent to the concerned doctor.
1.1 Objective
The main objective of my project is to develop a IOT based
anesthesia parameter monitoring machine and apply the
algorithms for prediction and make it in Low cost and User
Friendly. The device will also have safety features to ensure
smooth operation at the patient’s bedside and assist the
doctor. This procedure is easy, riskless and time saving for
doctors, patients and staff.
1.2 Challenges
The challenges faced while completing the project are
discussed briefly here. First everything started from the
decision of the domain. As, the project is in the health
domain the main issue with these types of domains is that a
lot of data is very restricted, closed and private. The case
study to find the base paper took a lot of hours and even
then, most of those are not open source and it struggled to
get the required data from the limited sources.
Next, the main concern arises when we need to get the
dataset for training the model. The anaesthesia parameter
dataset in the internet is either limited or can state it is
almost null available as open source. As Every hospital
maintains the data of patients as private no where we can
get the related data for the work. Anyway, after going
through a lot of research papers there is an experiment
performed by the University of Queensland on anaesthesia.
They collected the data from 32 patients during the surgery.
Finally, the dataset is taken from the experiment as the data
is open sourced.
Next, the selection of the sensors based on the parameters
needed for the doctor to get assistance. The parameters that
should be taken into consideration took a lot of time to

eliminate the unnecessary variables and get the required
ones. The step is sending the data from sensors to
thingspeak cloud through raspberry pi using python code.
There was an error while connecting the sensors and
sending the date. The code is changed andadjustedsuchthat
the data successfully reaches the thingspeak cloud. Again,
the main challenge is the project scope is not limited to one
device but is to develop for multiple devices. To do that the
website must be added with database connectivity and the
login for different accounts should be provided to the user.
2. LITERATURE SURVEY
2.1 ANESTHESIA CONTROL SYSTEM WITH MULTI
SENSOR USING ARDUINO IJIRAE: AM Publications,India,
2019
In the above paper, published 2019 by Thiyagarajan
proposed the concept of control system with multi sensors
using Arduino. The existing system is automatedandisdone
without the use of the doctor but here comes the clash. The
anaesthesia must be administered at different places on the
patient body based on the surgery. The paper only statesthe
offline use of the automated anaesthesia control but this is
different when we come to real time application as stated
above. Even the amount of anaesthesia should not be given
as if. It should be crossed checked with parameters and the
place where they have to be injected. So, This model fails
automatically.
2.2 Microfluidic Syringe Pump Using Arduino Dr.Azha
Periyasamy1, R. Jeya Kumar2, T. Karuppiah3 IJIRAEVol.
8, Issue 4, April 2019
In the above paper, published 2019 by Azha & co proposed
the concept of Microfluidic Syringe using Arduino. The
existing system is automated and is done without the use of
the doctor but here comes the clash. The anesthesia must be
administered at different places on the patient body based
on the surgery. The paper only states the offline use of the
automated anesthesia control but this is different when we
come to real time application as stated above. The
Microfluidic syringe cannot be used to do the automated
anesthesia machine control.
3. METHODOLOGY
3.1 BLOCK DIAGRAM
In the system, The Temperature sensor and humiditysensor
which is DHT11 is used and spo2 and heart rate senor is
MAX30102 is used and co2, wet sensors. All are taken as
parameters. Thesesensorsareconnectedtotheraspberrypi.
When the power supply is on the sensor send the data to the
raspberry pi and the data from the raspberry is sent to the
thinspeak cloud using inbuilt wi-fi module present in the
raspberry pi using API keysofthingspeak.Thetrainedmodel
is deployed in the website for the prediction andthechannel
data in thingspeak is exported to the website using APIkeys.
When the values cross thethresholdimmediatemessage will
be sent to the concerned person.
Figure 1: Block Diagram
3.2 REQUIREMENTS
3.2.1 HARDWARE:
1. RASPBERRY PI
2. DHT11 TEMPERATURE SENSOR
3. HEART RATE SENSOR
4. SPO2 SENSOR
5. CO2 SENSOR
6. WET SENSOR
3.2.2 SOFTWARE:
1. THINGSPEAK
2. TWILIO
3. WEBSITE, GOOGLE COLLAB

4. OUTPUTS AND DISCUSSION
Figure 2: ThingSpeak Channel
Figure 3: Website for results
Figure 4: Messages received to phone
5. CONCLUSION
This project attains high logistic accuracy and the as you see
the immediate message will be sent to the concerned staff
when the sensor crosses the threshold and even assist the
doctor with decision and the website stores the data asitgot
fetched from the ThingSpeak cloudandfurtherthisprojectis
not just limited to one device, the website supports for
multiple devices.
6. ACKNOWLEDGEMENT
This paper is a project work done by NAGA SAI RAM AELLA
under the guidance of DR. RADHA R, Associate Professor,
Department of Computer Science and Engineering, Vellore
Institute of Technology, as the final year project for the
fulfillment of the degree of Bachelor of Technology in
Computer Science and Engineering. I thank my guide for her
full support and encouragement throughout the project.
REFERENCES
[1] S. Dubey and A. Gambhir and S. K. Jain and A. V. Jha and
A. Jain and S. Sharma, “IoT application for the design of
digital drug administration interface” 2017
International Conference on Information,
Communication, Instrumentation and Control (ICICIC),
pp.1-5, 2017.
[2] ANESTHESIA CONTROL SYSTEM WITH MULTI SENSOR
USING ARDUINO IJIRAE: AM Publications,India, 2019
[3] John R L, Keith C H, Warren C R 2017 Low-cost
feedback-controlled syringe pressure pumps for
microfluidics applications. PLoS ONE 12(4): 1-12.
[4] David Miller, Sharon R Lewis, Michael W Pritchard,
Oliver J Schofield-Robinson, Cliff L Shelton, Phil
Alderson, et al., "Intravenous versus inhalational
maintenance of anaesthesia for postoperative cognitive
outcomes in elderly people undergoing non-cardiac
surgery", Cochrane Database Syst Rev, vol. 8, August
2018..
[5] Barkha Bindu, Ashish Bindra and Girija Rath,
"Temperature management under general
anesthesia:Compulsion or option", J Anaesthesiol Clin
Pharmacol, vol. 33, pp. 306-316, Jul-Sep 2017.

IOT Based Anesthesia Parameters Monitoring with Doctor Decision Assistance using Machine Learning

More Related Content

Similar to IOT Based Anesthesia Parameters Monitoring with Doctor Decision Assistance using Machine Learning (20)

More from IRJET Journal (20)

Recently uploaded (20)

IOT Based Anesthesia Parameters Monitoring with Doctor Decision Assistance using Machine Learning