Information Technology, in short IT, describes any technology that activates the storage, computation and accessing information within or outside the organization by authenticated users. The devices and technologies that come under IT are computers, mobiles, software’s, network, intranet, internet, websites, servers, telecommunications, Database Management Systems, Cloud Computing, Machine Learning, Artificial Engineering, R Programming etc., Currently, most of the organizations like Banks, Government sectors, Schools, Colleges, Universities, Healthcare sectors, Shopping malls, Companies etc. are using information systems and their applications for carrying out the general activities of storing, retrieving, sharing and analyzing information. Healthcare Sector is one important field which requires IT for implementing solutions for complex problems like X-Ray storing, collaborative sharing of patient disease details etc. for analysis and prediction. The reasons for using Cloud for the storage of health data is presented in a detailed fashion. In this paper, the application of Information Technology to Healthcare for predicting the death rate depending on few parameters is discussed. The programs for predictive analysis using Regression Models like Linear Regression and Multiple Regression are developed using R Programming. By predicting the value of death rate, measures could be taken to minimize it by taking necessary actions improving the doctor, hospital, medical facilities.

Introduction

As per the literature, Information Technology (IT) is the application of computers and internet to store, study, retrieve, transmit, and manipulate data, or information, often in the context of a business or other enterprise. IT is considered a subset of Information and Communication Technology (ICT). Presently, everyone is using different IT applications for accomplishing their daily activities. The IT applications include bank transactions, online recharge, e-commerce applications, bill payments, shopping, funds transfer, communication via the websites and social media apps, consumer health IT applications, mobile applications etc., The diverse domains where IT can be used are business, healthcare, science and engineering, education, fishing fields, Home, Departmental stores, markets, office buildings, traffic and transportations, Factories, Farms, Agriculture, weather Reporting Departments, Data centers, outer space etc.. The role of IT is very crucial in the above-specified areas. For example, in the education system, the details of students, staff, and the daily activities are stored in database systems and can be retrieved later. For better understandability of concepts, teachers are also using different presentation technologies like PPT, animations, audio, video, images etc., for delivering their lectures effectively. The weather reporting departments capture temperature using sensors and stores this data in computers for reporting purpose. Similarly, all other areas also use IT for performing their activities in a more effective manner.

IT in healthcare

Cloud Computing in Healthcare Sector

Healthcare is the maintenance and improvement of health using the activities of diagnosis, taking treatment, and following preventive measures. Healthcare is a very important aspect of our daily lives. Various information technologies are attached for secure storage of health data and for better analysis on decision making for the health care. Health sector industries like hospitals, pharmacy companies generate a huge amount of data to store the patient details, doctor details, disease details, X-rays details, the medicine details etc… Storing this data in local systems is difficult and requires the purchase of huge storage servers which is a costly process. An economic solution to store the data is by using an Internet Cloud. It is one of the innovative technologies which stores health data in remote servers in a secure way where an authenticated user can access the data in an efficient and effective way for better analysis. The reasons behind the healthcare industry for using cloud computing is based on the following key objectives:

- Unlimited Cloud Storage-Cloud provides a huge amount of storage space to the users based on requirements. Users need not buy costly servers for data storage
- Secure Cloud Platform- The users will access the cloud platform if they are having proper credentials for authentication purpose. If the credentials are valid then only they are able to access the data or computation services.
- Confidentiality- Cloud providers store data securely in Cloud with proper encryption algorithms like AES, ABE, CP-ABE etc.,
- Collaborative sharing-Cloud Computing enables the researchers for working on the same dataset for better analysis and thereby finding out better solutions for the problems.

Cloud provides various services to the users through the internet on a rental basis. The cloud services are mainly categorized into 3 types: infrastructure as a service (IaaS), software as a service (SaaS), and platform as a service (Paas). Users can access these services from anywhere any time through pervasive devices with a simple user interface and internet. For example, Physicians are targeting on breast and ovarian cancer for their research where huge massive amounts of information that is more than 2,000 DNA sequences are gathered from the Icahn School of Medicine at Mount Sinai. The data set is gigantic, more than 100 TB, which is difficult to be stored in a local system. The solution is to use a secure cloud-based platform for storing this gigantic data. The data from the cloud can also be collaboratively accessed by researchers for their analysis using Amazon Web Services, Microsoft Azure, RedHat etc…

Predictive Analysis of Health data through R Programming

Predictive analysis is a collection of statistical techniques which examines current and historical data to make predictions about future or otherwise unknown events. R is a programming language and software environment for statistical analysis and graphics. The approaches and techniques used to conduct predictive analytics can broadly be grouped into regression analysis techniques and machine learning techniques.

Regression analysis

Regression analysis (RA) is a statistical method for establishing relationships among variables. For modeling and analyzing various variables RA includes several techniques. The objective of regression analysis is to develop a relationship between dependent variable/response variable and the independent/ predictor variable. Depending on the context, there exists a variety of regression models that can be applied for performing predictive analysis. A few of the Regression Models are Linear regression model, Multi Linear Regression Model, Discrete choice model, Logistic regression model, Multinomial logistic regression etc..

Machine learning techniques

Machine learning, a division of artificial intelligence, is a process of developing systems that learn from the data available. Currently, Machine Learning involves several statistical methods for regression and classification. A few of the application fields of machine learning are credit card fraud detection, estimations, face and speech recognition, medical diagnostics etc., A few of the machine learning techniques are Fuzzy Inference Systems, Neural networks, Support vector machines (SVM), k-nearest neighbor, Geospatial predictive modeling etc..

Linear Regression

The most frequently used predictive analysis technique is Linear Regression. Regression estimates are used to describe data and to explain the relationship between dependent variable and independent variable.

The formula of linear regression is:

y = ax + b

where,

y — Dependent variable

x —Independent variable

a — Regression Coefficient

b — Constant

The linear regression model contains three stages –

- Analyzing the correlation and directionality of the data
- Establishing the model, i.e., fitting the line
- Evaluating the validity and usefulness of the model.

The uses of regression analysis are Normal analysis, Forecasting effect and Trend Forecasting.

**Normal analysis :**It is used to identify the strength of independent and dependent variable. For example, the relationship between death rate and doctor availability in health dataset.**Forecast effect :**It indicates how dependent variable is varied based on changing of the independent variable. For example, how death rate is varied as the doctor availability varies.- Thirdly, regression analysis predicts trends and future values. What will be the death rate if doctor availability s given?
- The following sample Health data set in Table 1 is considered for performing the predictive analysis with y, x1, x2, x3 and x4 as attributes.Table 1: Health Dataset

Y x1 x2 x3 x4 62 151 284 9.1000 109 78 130 433 8.699 144 70 140 739 7.1999 113 43 170 1792 8.89 97 The dataset details are as follows

y — Death rate per 1000 patients

x1 — doctor availability per 1000 patients

x2 —hospital availability per 1000 patients

x3 — annual per capita income of thousands of dollars

x4 — population density people per square mile.The requirement is to predict the death rate based on Doctor Availability. If the doctor availability is 179 what is the possible death rate?

For the dataset, the dependent variable is death rate (y) and the independent variable is doctor availability(x1).

The steps involved in the predictive analysis based on the linear regression is described below based on R programming.

- Carry out the analysis of gathering a dependent and independent variable values. In this, initially, death rate and doctor availability are gathered. These values are considered as a training set.
- Create a linear relationship model between death rate and doctor availability using the lm() function in R.
- Find the coefficients from the model created and create the mathematical equation.
- Get a summary of the relationship model to know the average error in prediction also called residuals.
- To predict the death rate based on doctor availability, use the predict() function in R.

R Program to Predict the Death Rate based on Doctor Availability

#independent vector

#Doctor availability per 1000 patients x < – c(151,130,140,170)

#The response vector- Death rate per 1000 patients

y <- c (62,78,70,43)

# Apply the lm() function. relation <- lm(y~x)

# Find death rate based on doctor availability

a <- data.frame(x=179) result <- predict(relation) print(result)Note: # line is a comment line

The result of the above program is 35.97389 i.e. if the doctor availability is 179 then there is a probability of 35.97389 death rates. By taking these statistics, proper measurements can be considered for reducing the death rate by increasing doctor availability.

In the above R program, x1 and y values are considered as vectors. lm() is a predefined function which provides the relationship between both response and predictor variables. A model summary is stored in relation variable. After completion of establishing a linear model, predict function is used to predict the death rate based on the doctor availability. The predict function takes doctor availability as input and generates possible death rate as output.

The above linear regression model is used when the response variable is based on a single predictor variable. Suppose, the death rate is based on the multiple predictors or independent variables linearly then multiple linear regression models is to be used.

Multiple Linear Regression

In Multiple Linear regression, the response variable is based on more than one predictor variables. In this dataset, if the death rate is based on doctor availability and hospital availability, then multiple linear regression is used. The formula for multiple linear regression is:y = a + b1x1 + b2x2 + …. + bnxn

where a, b1, … bn – coefficient variables

x1, x2, … xn—predictor /independent variables

y — response variableThe multiple regression models for health dataset in Table 1 is developed in R by just changing the lm() function used in the above program as given below

lm(deathrate~doctoravail+hospital availability, data=input)

The death rate is now predicted depending on Doctor availability and Hospital availability.

Conclusion

In this paper, the several applications of IT in daily life is discussed. The application of IT to Healthcare using the cloud is presented along with the objectives of using Cloud Computing for this sector. The basics of Predictive analysis using Regression Models and Machine Learning techniques is illustrated. Predictive analysis is carried out using the linear Regression and Multiple regression Models taking a sample Health dataset. The programs to predict the death rate is implemented in R Programming with built-in lm() and predict() functions. By predicting the value of death rate, measures could be taken to minimize the death rate by taking necessary actions.