IRJET - E-Health Chain and Anticipation of Future Disease

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 3066
E-health Chain and Anticipation of Future Disease
Rohit Dhonde1, Pradnesh Khedekar2, Pradeep Kshirsagar3 and Prof. Manasi Kulkarni4
1,2,3Member, Pillai College of Engineering, New Panvel, Maharashtra – 410206, India
4Guide, Pillai College of Engineering, New Panvel, Maharashtra – 410206, India
---------------------------------------------------------------------***----------------------------------------------------------------------
Abstract - E-Health Chain & AnticipatingFutureDiseasesis
a system which aims at maintainingElectronicHealthRecords
(EHRs) in a more efficient way as comparedtotraditional way
of storing and maintaining paper based health records. The
Digital prescription module can be used by patients to buy
medicines from pharmaceutical stores by just providing a
unique ID of the patient which helps pharmacists to accessthe
latest prescribed medicines. Electronic HealthRecords(EHRs)
allows doctors to access a patient’shealth recordssimplyfrom
one electronic file, doctors can browse, check results as they
are entered, together with image files like X-rays even from
remote hospitals. In an emergency scenario, a doctor can use
a patient’s Unique ID code to browse time-critical data, like
allergies, blood groups, recent treatments. In situations of
emergency, the historical data will assist the doctors to take
effective actions and safeguard the life of the patients. Tab
reminder alert helps patients to take medicines on time as
prescribed by doctor. System uses machine learning
algorithms to predict and examine futurediseaseswhichhelps
patients to take preventive measures.
Key Words: E-Health records(EHR), E-healthchain,
disease predictor, Random Forest, E-prescription, NLP,
E-ambulance
1. INTRODUCTION
E-Health Chain & Anticipating Future Diseases is a system
which aims at maintaining Electronic HealthRecords(EHRs)
in a more efficient way as compared to traditional way of
storing and maintaining paper based health records. Digital
prescription modules can be used by patients to buy
unique ID of the patient which helps pharmacists to access
the latest prescribed medicines. Electronic Health Records
(EHRs) allows doctors to access a patient’s health records
easily from a single electronic file, doctors can read test
results as they are entered, including image files such as X-
rays even from remote hospitals. E-Ambulance is a quick-
response solution that can detect and place a phone call for
the ambulance and send the ambulance to the required
destination. In an emergency situation, a doctor can use a
patient’s Unique code to read time-critical data, such
as,recent treatments, allergies, and blood type. In situations
of emergency, the historical data will assist the doctors to
take effective actions and safeguard the life of the patients.
Tab reminder alert helps patients to take medicines on time
as prescribed by doctor. System uses machine learning
algorithms to predict and examine future diseases which
helps patients to take preventive measures.
2. LITERATURE SURVEY
Heart Disease Prediction and Classification Using
Machine Learning Algorithms Optimized by Particle
SwarmOptimizationandAntColonyOptimization[1].The
aim of this work was to compare algorithmswithalldifferent
performance measures using machine learning. All data was
pre-processed and used to take a glance at the prediction.
every worked higher in some things and worse in others. K-
Nearest Neighbour K-NN, and Random Forest RF and
Artificial Neural Network MLP are the models apparently to
work best inside the knowledge set used in this study.
Experimental results show that the improvement hybrid
approach can increase the predictive accuracy of medical
data sets. The projected ways that are compared to
supervised algorithms based on existing approximate sets
and classification accuracy measurements are used to
measure the performance of the proposed approaches.
Therefore, the analysis section clearly incontestable the
effectiveness of hybrid PSO and ACO approaches to malady
diagnosing compared to different existing approaches. The
projected optimized model byFCBF,PSOandACOsucceedan
accuracy score of 99.65% with KNN and 99.6% with RF.
Liver disease prediction by using different decision tree
techniques[2]. The study used some decision tree formula
like J48, LMT, Random Forest, Random tree, REPTree,
decision Stump and Hoeffding Tree to predict the disease at
an earlier stage. These formulas provide varied results
supported Accuracy, Mean Absolute Error, Precision, Recall,
kappa statistics and Runtime. These techniques were
evaluated and their performance was compared. From the
analysis, Decision Stump outperforms well than different
algorithms and its achieved accuracy is 70.67%. The
performance measure used for comparison are listed within
the table (Table 2) the application of decision tree in
predicting disease can benefit in managing the health of
people. However, within the future, we are going to collect
the very recent data.
Disease prediction using machinelearning[3].Amachine
learning and new multimodal disease risk prediction
algorithmic rule supportedtheconvolutionalneuralnetwork
(CNN-MDRP) using structured and unstructured data. By
giving the input of symptoms we'll get correct disease
prediction as output, whichcanfacilitateusperceivethelevel
of disease risk prediction. This technique leads to low time

consumption and minimal value possible for illness
prediction.
Prediction of probability of disease based on symptoms
using machinelearningalgorithms[4].Amachinelearning
and new convolutional neural network based multimodal
disease risk prediction (CNN-MDRP) algorithm using
structured and unstructured data from hospital for effective
prediction of diseases. Existing work is not focused on both
data types in the area of healthcare. Compared to several
typical prediction algorithms, the proposed algorithm
accuracy prediction reaches 94.8% than that of the CNN-
based unimodal disease risk prediction (CNN-UDRP)
algorithm.
Performance Analysis of Machine Learning Algorithms
on Diabetes Dataset using Big Data Analytics[5]. Today’s
world people are more involved in their hectic schedules by
not taking care of their health, that ends up in chronic issues
like diabetics. in this paper, the author tries to provide a
comprehensive comparative study on different machine
learning algorithms. Thiscomparativestudyisdonebasedon
totally different metrics like Accuracy, Kappa, Precision,
Recall, Sensitivity and Specificity. The achieved results show
that RF formula is predicting hectic a lot of properly and
accurately.
Disease Prediction by Machine Learning Over Big Data
From Healthcare Communities[6]. A new convolutional
neural network based multimodal disease risk prediction
(CNN-MDRP) algorithm using structured and unstructured
data from hospital. To the best of our knowledge, none of the
existing work focused on both data types in the area of
medical big data analytics. Compared to several typical
prediction algorithms, the prediction accuracy of our
proposed algorithmreaches94.8%withaconvergencespeed
which is faster than that of the CNN-based unimodal disease
risk prediction (CNN-UDRP) algorithm.
Predictive Analytics for Chronic Kidney Disease Using
Machine Learning Techniques[7]. The predictive models
by using machine learning methods including K-nearest
neighbors (KNN), support vector machine (SVM), logistic
regression (LR), and decision tree classifiers to predict
chronic kidney disease.
Applying Machine Learning Techniques for Predicting
the Risk of Chronic Kidney Disease[8]. Data mining
techniques for various analysis of medical data may be a
smart methodology. The performance of decision tree
methodology was found to be 91 accurate comparedtonaive
bayes methodology. Classification algorithmic rule on
diabetes dataset performance was obtained as 94%
Specificity and 95% Sensitivity. We additionally found that
mining helps to retrieve correlations from attributes that
aren't direct indicators of the category that we tend to try to
predict. It is more working on enhancing the performance of
prediction system accuracy in neural networks and bunch
algorithmic rule knowledge analysis.
Data Analysis on Health Management Systems for
Improving Doctor's Advice on Patients[9]. Focusing on
improving doctors’ advice forpatientsviathedataofphysical
examinations, and helps patients get to know their physical
condition as accurately as possible. On basic data collecting,
we consider that the exchange of data with EMR is an
advisable way to let doctors obtain moreinformationabouta
patient's medical history. As an assisted diagnostic method,
the relationships about someunknown abnormal items with
special diseases can be discovered by analyzing the data
based on HMS.
Machine learning applications in cancer prognosis and
prediction[10]. The concepts of ML while we outlined their
application in cancer prediction/prognosis. Most of the
studies that have been proposed in the last few years and
focus on the development of predictive models using
supervised ML methods andclassification algorithmsaiming
to predict valid disease outcomes. Based on the analysis of
their results, it is evident that the integration of
multidimensional heterogeneous data, combined with the
application of different techniques for feature selection and
classification can providepromising toolsforinferenceinthe
cancer domain.
3. PROPOSED SYSTEM
Existing System Architecture
The existing manual method of maintaininga patient record,
maintaining doctor’s data,daytodayactivitiesandrequestis
hard and thus a system or application which mightcomplete
these tasks in an exceedingly straightforward to use is what
we are able to deliver through this application.
A health system consists of all organizations, folks and
actions whose primary intent is to market, restore or
maintain health. This includes efforts to influence
determinants of health in addition as additional direct
health-improving activities. A health system is thus over the
pyramid of publically closely-held facilities that deliver
personal health services. It includes,asanexample,a mother
caring for a sick kid at home; personal providers; behaviour
modification programmes; vector-control campaigns;
insurance organizations; activity health and safety
legislation. It includes inter-sectoral action by health
workers, as an example, encouraging the ministry of
education to market feminine education, a documented
determinant of higher health.
Proposed System Architecture
E-Health Chain & Anticipating Future Diseases is a system
which aims at maintaining ElectronicHealthRecords(EHRs)
in a more efficient way as compared to traditional way of
storing and maintaining paper based health records. The

Digital prescription module can be used by patients to buy
unique ID of the patient which helps pharmacists to access
the latest prescribed medicines. Electronic Health Records
easily from a single electronic file, doctors can read test
results as they are entered, including image files such as X-
rays even from remote hospitals. E-Ambulance may be a
quick-response answer which will notice and place a phone
call for the ambulance and quickly send the emergency
ambulance to the required destination. In an emergency
scenario, a doctor will use a patient’s unique ID code to scan
time-critical info, like allergies,blood type, and up to date
treatments. In situations of emergency, the historical data
will assist the doctors to take effective actionsandsafeguard
the life of the patients. A system in which it will store all the
patient’s historical data and the data will be used to predict
and examine the future diseases using Machine Learning
Algorithm.If user’s symptoms don't specifically match any
sickness within the info, then it's shows the diseases user
might most likely have based on his/hersymptoms..Thiscan
be used by Doctor to see Users historical data as well as by
the Pharmacist to provide medicine prescribed by Doctor’s.
User can login using Unique ID which canbeusedbyDoctors
to Access their previous records.Electronic Health Records
simply from one electronic file, doctors will browse take a
look at results as they're entered, together with image files
like X-rays even from remote hospitals.InE-ambulanceitcan
used to enter symptoms during the time when the patient is
in the ambulance.
Patient Registration: If Patient could be a new user he can
enter his personal details and he can use user Id and
password through that he will login to the system.
Patient Login:If Patient already has an existing account then
he/she will log into the system.
View Details: Patient and Doctor each will readtheir entered
details. Doctor can also read Patientsdetailsandpatients can
read solely doctors very little information.
Diseases Prediction: Patient can specify the symptoms
caused because of his unhealthiness. System will ask certain
questions regarding his illness and system predict. The
disease based on the symptoms specified by the patient and
system will also suggest doctor based on the disease.
Doctor: Doctor will record new data.
Pharmacist: Will provide medicine to the Patient which is
prescribed by the Doctor.
Role Based Access Control: Role-based access control
(RBAC) is a method of restricting network access based on
the roles of individual users within an enterprise. RBAC lets
Doctor’s have access rights onlytotheinformationthey need
to do their jobs(eg. Patient and Disease Profile)andprevents
them from accessing information that doesn't pertain to
them.
Figure1: Proposed system architecture
Preprocessing of data:
1. Punctuation Removal: The punctuationmarksare
removed from the text because they add no
meaning to the data thus of no use.
2. Blank/White space Removal: To remove leading
and ending spaces, you can use the strip() function.
It is helping to reduce the memory uses and
increase the efficiency of the model.
3. Stemming/Lemmatization: The aim of
lemmatization, like stemming, is to reduce
inflectional forms to a common base form. As
oppositionstemming,lemmatizationdoesn'tmerely
lop off inflections. Insteadit useslexical information
bases to urge the right base types of words.
4. Stop-Word Removal: The removal of Stopwords
(such as “is”,”the”.etc) is called Stop Word Removal.
Stopwords add very little meaningsoifremovedthe
database space is saved and processing speed
improves.
Disease Prediction Algorithm
Random Forest:
Random forests or random decision forests or associate
ensemble learning methodologyforclassification,regression
and different tasks that operates by constructing a large
number of decision trees at training time and outputting the
class that's the mode of the classes (classification) or mean
prediction (regression) of the individual trees. Random
decision forests correct for decision treeshabitofoverfitting
to their training set. In decision prediction the Random

Forest algorithm is used to process the symptoms as input
and gives diseases as output. Based on the probability of
predicted diseases top 4 diseases with brief information are
shown as output.
1) Preliminaries: decision tree learning
2) Bagging
The training algorithm for random forests applies the
general technique of bootstrap aggregating, or bagging, to
tree learners. Given a training set X = x1, ..., xn with responses
Y = y1, ..., yn, bagging repeatedly (B times) selects a random
sample with replacement of the training set and fits trees to
these samples:
For b = 1, ..., B:
1. Sample, with replacement, n trainingexamplesfrom
X, Y; call these Xb, Yb.
2. Train a classification or regression tree fb on Xb, Yb.
Accuracy Parameter:
1. Recall
Recall actually calculates how many of the Actual
Positives our model captures through labeling it as Positive
(True Positive).Recall shall be the model metric we use to
select our best model when there is a high cost associated
with False Negative.
Recall = (True Positive)/(True Positive+False Negative)
2. Precision
Precision talks about how precise/accurate your
model is out of those predicted positive, how many of them
are actual positive. Precision is a goodmeasuretodetermine
when the costs of False Positive is high.
Precision = (True Positive)/(True Positive+False Positive)
3. Confusion Matrix
A Confusion matrix is a table that is often used to
describe the performance of a classificationmodel ona setof
test data for which the true values are known.
REFERENCES
[1] Machine learning applications in cancer prognosis and
prediction.
[2] Predictive Analytics for Chronic Kidney Disease Using
Machine Learning Techniques
[3] PerformanceAnalysisofMachineLearningAlgorithmson
Diabetes Dataset using Big Data
[4] Disease Prediction by Machine Learning Over Big Data
From Healthcare Communities.
[5] Applying MachineLearningTechniquesforPredicting the
Risk of Chronic Kidney Disease.
[6] PREDICTION OF PROBABILITY OF DISEASE BASED ON
SYMPTOMS USING MACHINE LEARNING ALGORITHM.
[7] Data Analysis on Health Management Systems for
Improving Doctor's Advice on Patients.
[8] Heart Disease Prediction and Classification Using
Machine Learning Algorithms Optimized by Particle Swarm
Optimization and Ant Colony Optimization.
[9] DISEASE PREDICTION BY USING MACHINE LEARNING.
[10] LIVER DISEASE PREDICTION BY USING DIFFERENT
DECISION TREE TECHNIQUES.

IRJET - E-Health Chain and Anticipation of Future Disease

More Related Content

What's hot (20)

Similar to IRJET - E-Health Chain and Anticipation of Future Disease (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET - E-Health Chain and Anticipation of Future Disease