Data mining techniques for verification of medicine contents related to cardiac problems
DATA MINING TECHNIQUES FOR VERIFICATION OF MEDICINE CONTENTS RELATED TO CARDIAC PROBLEMS Shaikh Abdul Hannan 1, Pravin Yannawar2, Dr. R.R. Manza 2 , Dr. R. J. Ramteke 3, 1Vivekanand College, Aurangabad, India ([email protected]) 2 Department of Computer Science and Information Technology, Dr. B.A.M.U., Aurangabad 3 Department of Computer Science and IT, Reader, NMU, Jalgaon ([email protected])
ABSTRACT:
symptoms you experience depend on the type and
multidisciplinary areas, the knowledge discovery
severity of your heart condition. Learn to recognize
and data mining (KDD) field has made significant
your symptoms and the situations that cause them.
progress in the past decade. The field is vibrant
Call your doctor if you begin to have new
with significant impact on a wide variety of
symptoms or if they become more frequent or
science, business and technology areas where it has
severe. The most common symptom is angina.
become necessary to deal with exponentially
Angina can be described as a discomfort,
increasing volume of data. In this paper the
heaviness, pressure, aching, burning, fullness,
association rules as a means of identifying
squeezing or painful feeling in your chest. It can be
relationships among sets of symptoms, side effects
and medicine, which can be used to evaluate trends
Medicine Name
and classify groups. This concept is applied on medicine database to identify diseases on the basis
of symptoms and by avoiding medicine side effects prepare prescription for patients. In this paper
more than 100 patients’ data have been collected
and output is verified by expert Doctor. As per
Table. 1: Medicine ID, Medicine name and
Doctors opinion this output is nearby their outputs.
Medicine side Effect
The outcome of the result shown here will be useful for Doctors to give appropriate medicine to
Side Effect Description
patient on the basis of symptoms [1][2][3]
KEYWORDS: Introduction: The scalability of mining algorithms
has become a major research topic. One approach
Table 2 : Side effect ID and its description
to the scalability problem is to run data mining
Disease Name
algorithm on a small subset of the data. This
strategy can yield useful approximate results in a
fraction of the time required to compute the exact
solution, thereby speeding up the mining process
by orders of magnitude[4][5][6][7][8][9][10]. Data
mining (sometimes called data or knowledge
discovery) is the process of analyzing data from
Table 3 Disease ID, Disease Name and
different perspectives and summarizing it into
Symptoms ID
useful information. Information that can be used to increase
Symptom Description
both[11][12][13][14]. Data mining software is one
of a number of analytical tools for analyzing data.
It allows users to analyze data from many different
dimensions or angles, categorize it, and summarize
Table. 4 Symptoms ID and Symptoms Description
Technically, data mining is the process of finding
correlations or patterns among dozens of fields in large relational databases. One of the reasons
behind maintaining any database is to enable the
user to find interesting patterns and trends in the
Table. 5 Diseases ID and Medicine ID Heart disease symptoms and medicine : Experimental Analysis : In experiment samples
Coronary artery disease, heart attack -- each type of
from 100 patients has been collected from Sahara
heart disease has different symptoms, although
Hospital, under the guidance of Dr. Abdul Jabbar
many heart problems have similar symptoms The
(MD Medicine). According to observation of
samples the information has been arranged in
Heart attack with chest pain
formats like, Medicine which consists of Medicine
Medicine Name
Id, Medicine Name and Medicine Side effect as
Iso sorbide dinitrate, Ramipril, nikorandil,
shown in table 1. Where as table 2 shows the side
trimetazidine, metoprolol, clopidogrel, atorvastatin,
effect ID and description about side effect. The
corresponding ID Number of side effect of table 2
Conclusion
has written in appropriate row of medicine ID and
This work is based on association rules of
Medicine name in table 1. Table 4 has given the
data mining technique for verification of medicine
information about symptoms description and its ID
contents related to the cardiac problems. In the
number. The Symptoms ID number has been listed
experimental work it has been already considered
out in the corresponding rows of disease ID and
in all the cases of patients that the physical
disease description in table 3. The last table 5 as
examination of all the patients is found normal. In
given the relationship among disease ID with the
the result the information generated gives the
suitable medicine by giving medicine ID number
relation about disease with medicine, disease with
along with disease ID number. In this way the
symptoms, patient details with medicine. The final
samples information has been arranged in various
results have been verified by Dr. Abdul Jabbar
tables. Each table consists of around 150 entries
(MD Medicine). He is satisfactory about the work
out of them few important has given in tables 1 to
and all results. Further he has suggested that the
work can extend in future by considering physical
examination along with current inputs to generate
patients information, symptoms of disease are
good results. Therefore according to his
taken as input by considering all physical
suggestions this work in future may be extended to
examination were normal. Based on given inputs
the appropriate medicines were suggested as
References:
prescription for patient in result. By applying set
1. Keon-Myung Lee, “Mining generalized fuzzy
of association rules to avoid such medicine which
are harmful to the patient the final prescription is
generated as output. In the experimental work near
Conference, IFSA World Congress and 20th
about 100 side effects, 75 medicines, 10 different
NAFIPS, Page(s):2977 - 2982 vol.5, 25-28 July
diseases few symptoms has stored in data base.
The results produced are in various forms like
2. De-Xing Wang; Xue-Gang Hu; Xiao-Ping Liu;
individual patient, all patient, disease and medicine
Hao Wang; Jun Guo, “Research on model of
information, disease with symptoms report, specific
association rules mining with added-newly measure
disease wise symptom report, all patients, symptom
and suggested medicine information, individual
3. Rastogi, R.; Kyuseok Shim, “Mining optimized
information etc. One sample output of individual
association rules with categorical and numeric
patient which consists of patient name, age,
symptoms, disease and list of suggested medicine.
Like this more than 5 different formats as used to
produce patient reports as output of this research
4. Yingjiu Li; Peng Ning; Wang, X.S.; Jajodia, S.,
work. All results are satisfactory and verified by
“Discovering calendar-based temporal association
Individual report of the patient : Jamkar Gangadhar KeshawraoAge
Reasoning, Page(s):111 – 118, 14-16 June, 2001.
5. Xunwei Zhou; Hong Bao, “Mining Double-
PE Y -->Y: YES, Physical Examination,
Connective Association Rules from Multiple
Aurangabad found Normal (* PE : Physical
Symptom Description
Engineering, Page(s):271 – 279, Volume 4, 12-14
List of Possible Disease
6. Ishibuchi, H.; Yamamoto, T.; Nakashima, T.,
“Determination of rule weights of fuzzy association
List of Suggested Medicine
Page(s):1555 – 1558, Volume 3, 2-5 Dec. 2001.
Trimetazidine, Ramipril, Iso sorbide dinitrate,
Specific Disease and Medicine Information