https://www.irjet.net/archives/V6/i5/IRJET-V6I5985.pdf

1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 05 | May 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 6984
Blood Cross Portal: Blood Deferral Donor using Machine Learning
Koushik H. D1, K. R. Sumana2, Sunil Kumar N3
1PG Student, Department of MCA, NIE, Mysuru, India
2Assistant Professor, Computer Science Division,NIE College, Karnataka, India
3Mentor, Intrella Technologies, Karnataka, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - People giving blood might be deliberate non-
compensated bloodbenefactorsorsubstitutioncontributors as
vital by an individual from their very own family or network.
WHO prescribes willful non-compensated blood givers over
substitution givers because of the levelofbloodsafetyfromthe
two gatherings? In spite of the fact that an individual can
deliberately choose to give blood, they might be precluded
from giving blood because of reasons relating to the
benefactors' safety and additionally beneficiary security,
which is essentially alluded to ascontributordeferral. Deferral
might be worldly delay or changeless rejection from giving
blood due to being associated or affirmed with having an
irresistible sickness, hematological ailment, orwhateverother
ailment that will either impact the wellbeing of blood or
influence contributors' own wellbeing. In this venture,
everything about the of blood giver, regardlessof whetherthey
are benefactor deferral or not with complete subtleties.
Key Words: Blood Donor, Deferral Donors, Blood
Donation, Machine Learning, KNN
1. INTRODUCTION
Human blood is pitiful, vital andmuchprominent.Habitually
our country requires around 4 crore units of blood. In any
case, out of which the accessibility of bloodis40lakhunitsof
blood. Blood segments are frequently utilized in dangerous
circumstances of seriously sick patients. Hemoglobin and
organ gift are the main critical commitment that the
individual can serve to our country. It isn't destructive for a
grown-up individual to give blood. The anatomy of the giver
can recover blood inside couple of days. Similarly, the
individual can give a portion of his/her organs like kidney,
eye, pancreas and so forth. On the off fortunate thatthegiver
is hereditarily coordinated to recipient. There are sanguine
fluid gatherings, for example, A, B, AB, and O. The individual
with O-bunch blood is the widespread contributor and the
singular with AB+ bunch is the all-inclusive recipient. There
are different blood donation centers around India none of
them offer the extent for a paramount connection among
contributor and beneficiary. In our venture we propose
another and productive approach to beat such situations.
Cloud based organizationscanimprovecritical inemergency
blood movement since they can enable central and speedy
access to supplier's data and zone from wherever and for all
will and bourn any devices. Since almost everyonepasseson
a mobile phone with him, it ensures minute region following
and correspondence. Utilizing GPS (Global Positioning
System), we discover contributors closer to the area from
where the solicitation is created. The blood searcher
demands the blood giver accessibility by closer region and
required blood bunch by means of SMS to mechanized blood
donation center unit. The Global Position System module
which is interfaced to microcontroller gets message from
individual who requires blood fluid. Controller scans for
required data in the information base. In the event that it
matches it comes back with OK message and a SMS alert is
sent to power part for getting further assistance for the
individuals who need blood.
2. Overall Approach
In this paper, we developed an application thatidentifiesthe
deferral blooddonorsthroughquestionnairesandcategorize
them according to the duration of the reasons for their
postponement and communicate them and encourage them
to proactively donate their blood.
Fig -1: Block Diagram of Proposed Work
Describes the Sequence of Steps in Recognition of Blood
Availability.
2.1 Feature Extraction
This application is an open portal. So, any user who chooses
to use this application has to fill the form, in turn it is used to
predict whether donor is an immediate donor or deferral

2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 05 | May 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 6985
donor. The features that are considered in determining
whether the donor is deferral or not, are mentioned below.
Here donors are asked about whether they are under
medication,immunizationarenot. Similarly,applicationasks
further about diseases like Malaria, Blood Pressure and
Jaundice are considered into account. In case of female
candidate, ladies who are undergoing Menstruation cycle,
breast reading and pregnancy cycle are considered. Donors
who have underwent surgery, underweight [<50kg],
hemoglobin count is less are the factors in question.
Fig -1: Flow chart of donor
2.2 Data Extraction
Once the user or donor completes filling the given form,
the data is then compared with old data set and we obtain
desired data. This application uses Three main algorithm
to process the given sets. The new data set is considered
as Test data set and it is compared with training data set
(old data set). Firstly, we input these data sets in three
forms KNN algorithm i.e.
 Euclidean distance, Manhattan distance,
Chebyshev distance
Among the three we consider Euclidean distance is the best,
considering its time taken to process the data and accuracy
compared to other forms. Similarly, at same time we
compare test data and train data sets using SVM algorithm.
Here the data are grouped into clusters (Subsets) and gives
the output. The donors attribute is broken down into small
clusters and they are matched with previous ones. Finally,
we use K-Means to determine to deferral donors by picking
one of them as centroid and compare them with other
deferral donors and eliminate the less effected donor with
higher one. This works in iteration process until you can’t
find a centroid or left with only two deferral donors.
2.3 Classification of Donors
Once the process is done, the people who are eligible or
not, to donate blood is classified into two categories:
1. Deferral Donors – These donorsarenot eligiblefor
immediate donating of blood are called deferral
donors. These donors come under any of these
attributes mentioned below. Whether it maybeone
or several, they are deferral.
Table -1: Attributes taken in project
Attributes Result Duration classification
Medication yes After 3
Days
Deferral
Immunization yes After 3
Days
Deferral
Blood Pressure
[less than 90
and more than
180]
yes After 15
Days
Deferral
Jaundice yes After 15
Days
Deferral
Malaria yes After 15
Days
Deferral
Mensuration yes After 15
Days
Deferral
Pregnant yes After 15
Days
Deferral
Surgery yes After 90
Days
Deferral
Weight
[below 50]
yes After 90
Days
Deferral
Hemoglobin
Count
yes After 90
Days
Deferral
Breast Feeding yes After 150
Days
Deferral
2. Non-Deferral Donors – These donors are eligible
for immediate blood donation. These donorssatisfy
all the standards or norms set in ourdatabase(train
datasets).

3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 05 | May 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 6986
2.4 Result Analysis
Chart -1: Data Representation in Pie Chart
People Who can Donate after period of 3 months
People Who can Donate after 15 days
Deferral donors
People Who can Donate after 3 days
Donors
The overall deferral and non-deferral donorarerepresented
in pie graph.
Chart-2: Donors Fitness representation in Bar Graph
From the 1500 data sets taken after algorithm
implementation we can say 92.53% are donors fit to donate
in k-means and 95.308 in Support vector machine.
3. CONCLUSION
With rapid increase in accidents and donating of blood
frequently for money without knowing that it causes
temporal postponement or permanent exclusion from
donating. To bring awareness to peopleandencouragethem
to denote blood and to prevent people from donating blood
for money, and enhancement ofthehospitalityindustry.This
web application provides every details of the blood donor,
whether they are donor or not with complete details. With
this system, maintain records and classification of donors
has never been faster or easier. Thus, Blood Cross Portal:
Blood Deferral Donor Using Machine Learning) application
proves to be beneficial to donors, hospitals and needy
people.
ACKNOWLEDGEMENT
I have to thank my guide K. R. Sumana for enormousheading
and Sunil Kumar N for your tutoring through this
undertaking. My real thankfulness to watchmen for their
endowments and accomplices for their assistance.
REFERENCES
[1] Bravo, M., Kamel, H., Custer, B., Tomasulo, P. “Factors
associated with fainting – before, during and after whole
blood donation,” Vox Sanguinis (2011) 101, 303– 312.
[2] Eder, Anne F.; Notari IV, Edward P.; and Dodd, Roger Y.
“Do reactions after whole blooddonationpredictsyncopeon
return donation?” TRANSFUSION Volume 52, December
2012.
[3] Erraguntla, M., Gopal, B., Ramachandran, S., Mayer, R. J.
2012, “Inference of Missing ICD9 Codes Using Text Mining
and Nearest Neighbor Techniques,” Hawaii International
Conference on System Sciences, 2012.
[4] Kamel, Hany;Tomasulo,Peter;Bravo,Marjorie; Wiltbank,
Thomas; Cusick, Robin; James, R. C; and Custer, Brian.
“Delayed adverse reactions to blood donation,”
TRANSFUSION, Volume 50, 556-565.
[5] Kamel, Hany; Townsend, Mary; Schroeder, Kadi; Bravo,
Marjorie; Erraguntla, Madhav; Whitaker, Barbee; Land,
Kevin; Tomasulo, Peter. “UNITED STATES DONOR
HEMOVIGILANCE SYSTEM: Experience of 3 participating
blood centers,” AABB Annual Meeting & CTTXPO 2011, San
Diego, CA.
[6] KBSI, Environment, Epidemiology, and Etiology
Surveillance and Analysis Toolkit Phase II, 2009, W81XWH-
08-C-0756.
[7]KBSI.2013"DonorHART™System,"
https://www.donorhemovigilance.org/Default.aspx
[8] Tomasulo, Peter; Kamel, Hany; Bravo, Marjorie; James,
Robert C.; and Custer, Brian. “Interventions to reduce the
vasovagal reaction rate in young whole blood donors,”
TRANSFUSION, Volume 51, July 2011.