More Related Content
Similar to Fast and effective heart attack prediction system using non linear cellular automata
Similar to Fast and effective heart attack prediction system using non linear cellular automata (20)
More from IAEME Publication
More from IAEME Publication (20)
Fast and effective heart attack prediction system using non linear cellular automata
- 1. International Journal of Computer and Technology (IJCET), ISSN 0976 – 6367(Print),
International Journal of Computer Engineering Engineering
and Technology (IJCET), ISSN 0976May - June Print) © IAEME
ISSN 0976 – 6375(Online) Volume 1, Number 1,
– 6367( (2010),
ISSN 0976 – 6375(Online) Volume 1
IJCET
Number 1, May - June (2010), pp. 196-206 ©IAEME
© IAEME, http://www.iaeme.com/ijcet.html
FAST AND EFFECTIVE HEART ATTACK PREDICTION
SYSTEM USING NON LINEAR CELLULAR AUTOMATA
N.S.S.S.N Usha devi
Post Graduate Student of C.S.E
University College of Engineering, JNTU Kakinada
E-mail: usha_nedunuri@yahoo.com
L.Sumalatha
Head, Department of C.S.E
University College of Engineering, JNTU Kakinada
E-mail: ls.cse.kkd@jntukakinada.edu.in
ABSTRACT
These days the Cellular Automata based Classifier have been widely used as tool
for solving many decisions modeling problems. Medical diagnosis is an important but
complicated task that should be performed accurately and efficiently and its automation
would be very useful. A system for automated medical diagnosis would enhance medical
care and reduce costs. In this paper have proposed a Cellular Automata Classifier, Non
Linear Fuzzy Multiple Attractor Cellular Automata (NNFMACA) for the prediction of
Heart attack. A set of experiments was performed on a sample database of 5000 patients’
records, 13 input variables (Age, Blood Pressure, Angiography’s report etc.) are used for
training and testing of the Cellular Automata Classifier. The performances of the
NNFMACA were evaluated in terms of training performances and classification
accuracies and the results showed that the proposed NNFMACA model has great
potential in predicting the heart disease.
Keywords: Cellular Automata, Data Sets, Heart Attack, NNFMACA
I. INTRODUCTION
Clinical decisions are often made based on doctor’s intuition and experience
rather than on the knowledge rich data hidden in the database. This practice leads to
unwanted biases, errors and excessive medical costs which affects the quality of service
196
- 2. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
provided to patients. A major challenge facing healthcare organizations (hospitals,
medical centres) is the provision of quality services at affordable costs. Quality service
implies diagnosing patients correctly and administering treatments that are effective. Poor
clinical decisions can lead to disastrous consequences which are therefore unacceptable.
Hospitals must also minimize the cost of clinical tests. A majority of areas related to
medical services such as prediction of effectiveness of surgical procedures, medical tests,
medication, and the discovery of relationships among clinical and diagnosis data also
make use of Data Mining methodologies . Proffering valuable services at reasonable
costs is a chief confront envisaged by the healthcare organizations (hospitals, medical
centres). Valuable quality service refers to the precise diagnosis of patients and proffering
effective treatment. Poor clinical decisions may result in catastrophes and so are not
entertained. It is also necessary that the hospitals reduce the cost of clinical test. This can
be attained by the making use of proper computer-based information and/or decision
support systems. Prevention of HD can be approached in many ways including health
promotion campaigns, specific protection strategies, life style modification programs,
early detection and good control of risk factors and constant vigilance of emerging risk
factors.
II. DESCRIPTION OF DATABASE
The heart-disease data base at Meenakshi Medical College, Kanchipuram consists
of 500 cases where the disorder is one of four types of heart-disease or its absence.
III. RELATED WORKS
A novel technique to develop the multi-parametric feature with linear and
nonlinear characteristics of HRV (Heart Rate Variability) was proposed by Heon Gyu
Lee et al.. Statistical and classification techniques were utilized to develop the multi-
parametric feature of HRV. Besides, they have assessed the linear and the non-linear
properties of HRV for three recumbent positions, to be precise the supine, left lateral and
right lateral position. Numerous experiments were conducted by them on linear and
nonlinear characteristics of HRV indices to assess several classifiers, e.g., Bayesian
197
- 3. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
classifiers, CMAR (Classification based on Multiple Association Rules), C4.5 (Decision
Tree) and SVM (Support Vector Machine). SVM surmounted the other classifiers.
A model Intelligent Heart Disease Prediction System (IHDPS) built with the aid
of data mining techniques like Decision Trees, Naïve Bayes and Neural Network was
proposed by Sellappan Palaniappan et al.. The results illustrated the peculiar strength of
each of the methodologies in comprehending the objectives of the specified mining
objectives. IHDPS was capable of answering queries that the conventional decision
support systems were not able to. It facilitated the establishment of vital knowledge, e.g.
patterns, relationships amid medical factors connected with heart disease. IHDPS subsists
well being web-based, user-friendly, scalable, reliable and expandable.
IV. CELLULAR AUTOMATA
4.1CELLULAR AUTOMATA (CA) AND FUZZY CELLULAR
AUTOMATA (FCA)
A CA[6],[8] , consists of a number of cells organized in the form of a lattice. It
evolves in discrete space and time. The next state of a cell depends on its own state and the
states of its neighbouring cells. In a 3-neighborhood dependency, the next state qi (t + 1) of a
cell is assumed to be dependent only on itself and on its two neighbours (left and right), and is
denoted as
qi(t + 1) = f (qi−1(t), qi(t), qi+1(t))
th th
Where qi (t) represents the state of the i cell at t instant of time, f is the next
state function and referred to as the rule of the automata. The decimal equivalent of the
next state function, as introduced by Wolfram, is the rule number of the CA cell. In a 2-
state 3-neighborhood CA, there are total 256 distinct next state functions.
4.2 FCA FUNDAMENTALS
FCA [2], [6] is a linear array of cells which evolves in time. Each cell of the array
assumes a state qi, a rational value in the interval [0, 1] (fuzzy states) and changes its
state according to a local evolution function on its own state and the states of its two
neighbours. The degree to which a cell is in fuzzy states 1 and 0 can be calculated with
198
- 4. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
the membership functions. This gives more accuracy in finding the coding regions. In a
FCA, the conventional Boolean functions are AND , OR, NOT.
V NFMACA BASED PATTERN CLASSIFIER
NFMACA [13] classifies a given set of patterns into k distinct classes, each class
containing the set of states in the attractor basin. A NFMACA is a special class of FCA
that can efficiently model an associative memory to perform pattern recognition
classification task. Its state transition behaviour consists of multiple components - each
component, as noted in Figure 1, is an inverted tree, each rooted on a cyclic state. A cycle
in a component is referred to as an attractor. In the rest of the paper we consider only the
NFMACA having the node with self loop as an attractor state. The states in the tree
rooted on an attractor form an attractor basin.
Figure1 Inverted tree
EXAMPLE 1:
Let us have two pattern sets S1 ={(0.00,0.00, 0.25), (0.00, 0.25, 0.00), (0.25, 0.25,
0.00), (0.00,0.50, 0.00), (0.00, 0.00, 0.00), (0.25, 0.00, 0.00), (0.50,0.00, 0.00), (0.00,
0.00, 0.25), (0.00, 0.00, 0.75), (0.00,0.50,0.25)} (Class I)
S2 = {(0.75, 1.00, 0.00), (1.00,0.75, 0.50), (1.00, 1.00, 1.00), (0.75, 1.00,
1.00),(1.00,1.00, 0.75), (1.00, 0.75, 1.00), (0.50, 0.75, 1.00), (1.00,0.75, 0.75), (0.75,
1.00, 0.75), (0.75, 0.75, 1.00)} (Class II) with three attributes.
In order to classify these two pattern sets into two distinct classes, Class I and II
respectively, we have to design a NFMACA such that the patterns of each class falls in
distinct attractor basins.
The basins have certain properties depending on the input loaded, it will go to
autonomous state and it gives the result. When the NFMACA is loaded with an input
199
- 5. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
pattern say P = (1.00, 0.50, 0.00) and is allowed to run in autonomous mode, it travels
through a number of transient states and ultimately reaches an attractor state (0.50, 0.50,
0.00) the attractor representing Class II. Here (0.00, 0.25, 0.00), (0.50, 0.50, 0.00) are
attractor basins named b, d respectively.
5.1 NFMACA BASED TREE-STRUCTURED CLASSIFIER
Like decision tree classifiers, NFMACA based tree structured classifier
recursively partitions the training set to get nodes (attractors of a NFMACA) belonging to
a single class. Each node (attractor basin) of the tree is either a leaf indicating a class; or a
decision (intermediate) node which specifies a test on a single NFMACA.
Suppose, we want to design a NFMACA based pattern classifier to classify a
training set S = {S1, S2, · , SK} into K classes. First, a NFMACA with k-attractor basins is
generated. The training set S is then distributed into k attractor basins (nodes). Let, S’ be
the set of elements in an attractor basin. If S’ belongs to only one class, then label that
attractor basin for that class. Otherwise, this process is repeated recursively for each
attractor basin (node) until all the examples in each attractor basin belong to one class.
Tree construction is reported in. The above discussions have been formalized in the
following algorithm. We are using genetic algorithm classify the training set.
ALGORITHM 1: NFMACA TREE BUILDING
Input: Training set S = {S1, S2, · ·, SK} db sets
Output: NFMACA Tree with disjoint disease parameters.
Partition(S, K)
Step 1: Generate a NFMACA with k number of attractor basins with db.
Step 2: Distribute S into k attractor basins (nodes) with disease parameters.
Step 3: Evaluate the distribution of examples in each attractor basin (node).
Step 4: If all the examples (S’) of an attractor basin (node) belong to only one class, then
label the attractor basin (leaf node) for that class
Step 5: If examples (S’) of an attractor basin belong to K’ number of classes, then
Partition (S’, K’).
Step 6: Stop.
200
- 6. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
VI EXPERIMENTAL RESULTS
We have developed a Non Linear Fuzzy Multiple Attractor Cellular Automata
Classifier for both testing and training Figure 2-5. The interfaces and results are displayed
below.
6.1 HEART ATTACK PREDICTION
The design of the intelligent and effective heart attack prediction system with the
aid of CA network is presented in this section. The method primarily based on the
information collected from precedent experiences and from current circumstances, which
visualizes something as it may occur in future, is known as prediction. The degree of
success differs every day, in the process of problem solving on basis of prediction. CA
networks are one among the widely recognized Artificial Intelligence (AI) machine
learning models, and a great deal has already been written about them. A general
conviction is that the number of parameters in the network needs to be associated with
the number of data points and the expressive power of the network.
EXAMPLE:
If
Male And age < 30 And CA Smoking = Never And Overweight = No And Alcohol =
Never And Stress = No And High saturated fat diet (hsfd) = No And High salt diet (hsd)
= No And Exercise = CA normal And Sedentary Lifestyle (Inactivity) = No And
Hereditary = No And Bad Cholesterol = Low And NCA BLOOD Sugar = CA normal
And NCA BLOOD Pressure = CA normal And Heart Rate
= CA normal
Or
Male And age > 50 and age < 70 And Smoking = Current And Overweight = No
And Alcohol = Past And Stress = No And High saturated fat diet (hsfd) = No And High
salt diet (hsd) = Yes And Exercise = High And Sedentary Lifestyle (Inactivity) = No And
Hereditary = No And Bad Cholesterol = Low And NCA BLOOD Sugar = Normal And
NCA BLOOD Pressure = Normal And Heart Rate = Normal Then Risk Level = Normal
201
- 7. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
Figure 2 Training Interface
202
- 8. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
Figure 3 Target Vs Best Fit
203
- 9. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
Figure 4 Testing Interface
Figure 5 Testing Interface
204
- 10. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
VII. CONCLUSIONS
In this paper, we have presented Fast and effective heart attack prediction
methods using Non Linear Fuzzy Multiple Attractor Cellular Automata.. Firstly, we have
provided an efficient approach for the extraction of significant patterns from the heart
disease for the efficient prediction of heart attack ,Based on the calculated significant
weight age at the NFMACA tree, the patterns having value greater than a predefined
threshold were chosen for the valuable prediction of heart attack. Five goals are defined
based on business intelligence and data exploration. The goals are to be evaluated against
the trained models. . We also tested the proposed classifier with 30,000 real time data sets
and it was found very effective in predicting the heart attack.
VIII. ACKNOWLEDGMENT
I thank all the faculty members of department of C.S.E, University College of
Engineering, JNTU Kakinada for their valuable support during my project. I also thank
Dr J.V.R Murthy and Dr M.H.M Krishna Prasad for their valuable suggestions during my
project period. I also thank Dr K.Karnan, Chief Superintendent of Meenakshi Medical
College, Kanchipuram for providing the real time data sets. Finally I thank all my fellow
class mates for their consistent encouragement.
IX. REFERENCES
[1] Sellappan Palaniappan, Rafiah Awang, "Intelligent Heart Disease Prediction System
Using Data Mining Techniques", IJCSNS International Journal of Computer Science
and Network Security, Vol.8 No.8, August 2008.
[2] Franck Le Duff, Cristian Munteanb, Marc Cuggiaa, Philippe Mabob, "Predicting
Survival Causes After Out of Hospital Cardiac Arrest using Data
Mining Method", Studies in health technology and informatics, Vol. 107, No. Pt 2,
pp. 1256-9, 2004.
[3] Shantakumar B.Patil al. “Intelligent and Effective Heart Attack Prediction System
Using Data Mining and Artificial Neural Network”, European Journal of Scientific
Research, ISSN 1450-216X Vol.31 No.4 (2009), pp.642-656
205
- 11. International Journal of Computer Engineering and Technology (IJCET), ISSN 0976 – 6367(Print),
ISSN 0976 – 6375(Online) Volume 1, Number 1, May - June (2010), © IAEME
[4] Latha Parthiban and R.Subramanian,” Intelligent Heart Disease Prediction System
using CANFIS and Genetic Algorithm” in International Journal of Biological and
Life Sciences 3:3 2007,pp157-161.
[5] K.Srinivas et al. “Applications of Data Mining Techniques in Healthcare and
Prediction of Heart Attacks” (IJCSE) International Journal on Computer Science and
Engineering Vol. 02, No. 02, 2010, 250-255.
[6] Pradipta Maji, Samik Parua, Sumanta Das, and P. Pal Chaudhuri, Cellular Automata
in Protein Coding Region Identification, IEEE Proceedings of 2nd International
Conference on Intelligent Sensing and Information Processing (ICISIP-05), India, pp.
479--484, January 2005.
[7] Pradipta Maji, Biplab K. Sikdar, and P. Pal Chaudhuri, Cellular Automata Evolution
for Pattern Classification, Proceedings of 6th International Conference on Cellular
Automata for Research and Industry (ACRI-04), Netherland, pp. 660--669, October
2004.
206