Data mining has also been named as data "knowledge mining." Data mining is an important process, intelligent method is applied to extract data model. Data mining is a process for analyzing large amounts of data; it was found New and hidden information. In other words, it is the analysis of data from a different perspective; summarize its course for some useful information. The purpose of data mining is to discover novel database, interesting and useful information. By convention, the disadvantage data manually, which is to analyze the many hidden and potentially useful information cannot be Analysts understand. Today, many organizations, including modern hospitals are collecting large amounts of this data requires an automated way to extract useful information. Therefore, the medical field is an important area the application of data mining. Through data mining, the interesting knowledge can be extracted and applied to increase Work efficiency. In mining medical data highlights of this writing they were discussed. In this study, we focused on data mining techniques and importance of medical system for some of the problems mining information technology in the field of medicine, the focus is Development and implementation of a telemedicine solution to improve medical services to the patient.


Comparative study to evaluate the performance of data mining techniques to predictAnd the ability to diagnose breast cancer survive the disease has taken place, has been evaluated Mining techniques popular data such as decision trees, neural networks, recently k-Logistic regression algorithm several neighbors, the K-means algorithm, Apriori algorithm, Using data from breast cancer, SEER data set of partial least squares regression. The results of this study indicate that the partial least squares regression in maximum Data mining techniques to predict the feasibility and the diagnosis of breast cancer disease the techniques used in the study.

The field of medicine is an area where most clinical and biological researchers through his research, but with the development of the information age to develop new knowledge Technology have changed all health information rapidly. These researchers have paid special attention to their cancer Related research, since it is one of the most diseases caused by the death, and if they can To improve the lives of patients with cancer, or if they can provide any assistance to improve cancer Prevention, it would be worth the effort to be considered. As mentioned above, this study merge previous studies have been performed Data mining techniques used to predict the feasibility of different types of cancer. due to lack of Already in operation for cancer survivors forecast e arlier efforts in Sri Lanka In the context of Sri Lanka, this study overall survival of cancer care is provided By studying in a different background with different sets of data like the SEER(Surveillance, Epidemiology and End Results) data set(Lakshmi, 2013).

Traditionally, such as using the Kaplan-Meier and Cox proportional hazards (Cox, 1984) Traditional statistical techniques were used to model survival. These techniques it is based on conditional probability model, a survival probability estimates. With A new method of flow data from the areas of knowledge discovery and data mining progress has a existence. These methods proved more powerful than the Traditional statistical methods. Now, focus on research has been carried out alone Using data mining techniques to predict cancer survival. A study has been carried out using data mining techniques (three decision trees, neural networks and support SVM), with the most commonly used statistical analysis techniques Logistics Forecasting model of the evolution of prostate cancer survival regression. This dataset has There are about 120 000 records and 77 variables. A method of cross validation times KIt has been used to build the model, evaluation and comparison. The results show Support vector machine is more accurate predictions (and 92.85%, accuracy test)the domain name, followed by artificial neural networks and decision trees(William, 2009).

Another comparative study of two popular algorithms data mining (artificial Networks and decision trees), along with one of the most commonly used statistical methods Neural(Logistic Regression), the use of large data sets with breast cancer to develop predictive models Patients. The results showed that the decision tree (C5) with the best forecast of 93.6%Precision shows retention (this prediction accuracy is better than any reported Literature), artificial neural network after it became second with 91.2% regression model the logistic lively ACCU is the worst of the three, with a 89.2% accuracy(Delenet al., 2005).

The clinical study of its kind in the medical field, but progress Researchers IT trends in the field of health must Medical information. Data mining technology plays an important role in medical informatics. Especially when it comes to forecasting or prediction it is known as a disease of survival Medical prognosis, data mining plays an important role. Eventually medical prognosis becomes very It is important to raise patients, especially cancer patients who were diagnosed with morality. Although the prognostic significance of cancer is increasing, that is, in practice, a method for predicting.

Data mining reveals new acquaintances for biomedical and clinical care Decision has great potential to improve treatment and quality of hospital Increased patient survival. One of the applications of diagnostic accuracy Data mining tools to create a successful outcome. Data Mining intends to provide Current technology and knowledge discovery by the database system in the survey Data mining technology, the use of medical research today. to discuss Making the diagnosis and medical advances hidden patterns of disease Related databases available. The use of data mining to discover relationships such as Representation between health and disease is become very easy. The Committee also reviewed the It can be used for processing and data classification tool. WEKA has been using drugs Use brings to light most drug addicts began Age 20yrs.It is to let the pharmacist know all kinds of diseases are and I caused the drug in your life or long-term heavy intake. Therefore, in order to make the expert system this awake for youth at risk of drug use, but also to remind affected People (Kaur and Bawa, 2015).

Health care system faces several challenges, now, thanks changes Place on health systems worldwide. These challenges are mainly composed of representatives Economic constraints (the sharp increase in costs, financial problems), and in an increasingly stressed Accountability and transparency, which is the change made in the field of education, in fact, Biomedical research constantly increasing complexity of the particular concerns Investigation. This is a new partnership in the system of health care and the great proposal in the development of the IT industry it showed a significant change in pattern is taking place. This requires Focus on interaction, collaboration and exchange of information and knowledge increases, all these are likely to be new leader in health institutions information technology mining, in order to establish and maintain optimal health outcomes. Data mining is a domain now very important day because it provides advanced data analysis techniques to extract Day knowledge massive data collected and stored for each system Bases. Medical institutions data mining can provide valuable information for patient diagnosis and treatment planning, management of customer relations, the organization Resource management and fraud detection (GHEORGHE and PETRE, 2014).

Fig1. Architecture for integrating data mining techniques into a telemedicine system

Mining medical data related to the health field, which helps capture the potential areas,the relationship between the attributes. This knowledge is in the modern world has becoming  dispensable element Clinical data is very large, manual interpretation is impossible. Including transactional mining medical data base Tools have real-world data such as medical diagnosis, decision making expert as well. Medical Data Mining also contribute to evidence-based research is necessary for the industry health care. Although Studies in many areas of medical data mining, in this article, we focus on the mining method or the method to help find regular common diseases. It occurs in a fixed time interval pattern is considered to be-frequent. Periodic Newspaper frequent pattern are trend data often a fixed time interval. What mode but configuration shows an interesting article on facts? Typically, this Interesting expert help businesses make decisions data. The minimum support of filter aid It does not occur with a given frequency and pattern Periodic guidelines can help control the maximum time She appeared twice in a given pattern between the difference Data set. This medical data mining can bring Knowledge of the elements of interest rate was found (Khaleel et al., 2014).

In recent decades, the rapid changes in information technology have greatly changed Multimedia functions and activities. Data mining has become increasingly popular to extract knowledge from Sets of multimedia, such as a combination of audio, video, voice, text, Web pages, images and various types of data these data Set. This data is a major challenge manually removes the hidden embedded multimedia collections, useful knowledge, without the use of new Technology and powerful tools. This challenge drives the need to develop tools and techniques that can data mining can be used for the above types of data sets. Nowadays multimedia data is a common need in a variety of applications; Archives multimedia data requires large storage capacity. Multimedia data mining research is an interdisciplinary, generic data mining the theory and technology applied to sets of multimedia data, in order to promote specific knowledge multimedia Discovery task.

PROBLEM STATEMENT

Knowledge discovery problem is that modern medicine Produce almost every day, a large number of heterogeneous data. Those who handle these data to understand, there is a Widen gap between data collection and understanding between. The data mining techniques are used to help and address humanity this problem. From the collection of medical data in humans, it is the design of ethical, legal and social aspects Prevent patients from abuse and misuse of their data. These problems may be related to the ownership of the data, for fear of the procedures, Privacy and human data, the expected benefits and safety of administrative affairs. So this research point out these issues and some challenges of data mining techniques.

OBJECTIVES

The purpose of this study is to present and analyze a number of tools and data mining Better prediction of many medical conditions, giving Healthcare portion has a greater capacity and more efficient. In embodiment Research objectives of this study was to answer the following five questions:

       I.            If data mining provides an efficient way to extract the desired clinical from the original heterogeneous data information voluminous?

    II.            Is there any promising technology for predicting and forecasting internal diseases with high accuracy and low cost?

 III.            For a variety of issues on which, and the challenges in data mining Medical practice?

 IV.            What it is the study of drug abuse and related reasonable profit?

    V.            How can I get information simultaneously and minimize effort Expert systems, such as abuse of drugs and other medical problems?

METHODOLOGY

To complete this research there are different steps involved to clearly define the methodology adopted. In this research we use a methodology for their comparative study, comparison and some challenges of data mining techniques in medicines. As a part of the research design, it will be decided upon the procedure which wanted to adopt to collect the data. In this phase actually data will be collected. The data will be collected by commence interviews, conduct nominal/focus group discussions. The data will be collected by reading journals, survey papers and by internet. The way will be chosen to analyze the information that collected. The type of information will be qualitative or attitudinal. Writing the report is the last and the most difficult step of this research process. This report will tell the world what we have done, what we have discovered and what conclusions we have drawn from our findings. Then the conclusion will be written. The method used in this study will be the choice and through magazine Publications in computer science, engineering and healthcare. In previous studies, profiles and knowledge available, the relevant papers, working papers and conference papers are also included Investigation. Initially, this study are summarized studies have been conducted in the previous Using data mining techniques of care. This only takes into account the latest research this will be included in the last five to six years. Depending on the mode, the data mining techniques were used in these studies, the examples are classified together.