Health Care IT Infrastructure: a Software Engineering Perspective

Health Care IT infrastructure: A Software Engineering Perspective.

Samuel Kasimalla under Supervision of Dr Box for CSC 532 Term paper

ABSTRACT: Health care infrastructure is an important topic under general consideration. With many new diseases and outbreaks, health care always remains to be one of the biggest industries in terms of expenditure and market size. Health care is again such a field in which negligence can be very costly. This requires that all the implementation should be utmost reliable.

Software engineering comes into play when the reliability factor is needed to be considered. For a software to be reliable the efficiency, planning, execution should be perfectly engineered.

INTRODUCTION

The research problem chosen for the paper is loosely based on software engineering principles on distributed systems. More and more systems use a centralized data base with nodes in varied geographical locations. So we need high speed networks capable of transferring huge amount of data that is required by data miners. Maintaining the reliability is absolutely essential, but other factors should also be considered in the software engineering principles. This paper would touch upon different aspects of software engineering and management in the field of Health care. Some of them are

1. Planning and Implementation a project
2. Phases of implementing a health care based large project.
3. Frameworks
4. Cloud based approach
5. Data Mining in Health care.
6. Privacy preserving sharing of information
7. High performance computing for data mining

PLANNING AND IMPLEMENTATION A PROJECT

Hospitals and nations are allocating significant amount of resources for introducing new technology. “Cost accounting analysis is a multivariate function that includes determining the amount, based upon a strategic plan and financial resources, of funding to be allocated annually for medical equipment acquisition and replacement” [8].

The process of selecting and acquiring medical technology has not been well coordinated in most hospitals until recent times (Sprague, 1988). This is true with most countries. The focus of the thing should be to first identify its goals, secondly select and define priorities, and finally allocate there sources, although limited, with which we have to attain those goals [8]. During the initial planning process percentage of resources should be allocated instead of the actual values. The actual figures are subject to change with the time. Health care related projects can be expensive but there are always alternate ways to provide for them.

As given by [8]. A good planning package should

(i)Provides for guiding strategy for allocation of limited resources

(ii)Maximizes the value provided by resources invested in medical technology

(iii)Identifies and evaluates technological opportunities or threats

(iv)Optimizes prioritization in systems integration, facility preparation and staff planning

(v)Meets or exceeds standards of care

(vi)Reduces operating costs

(vii)Reduces risk exposure.

There exists a relationship between the methods and information. Methods help to take decisions regarding the management of the medical technology. Medical technology is used in the complex environment of the healthcare delivery system, which includes the variances among users, applications and cultures from one hospital to another [8]

PHASES OF IMPLEMENTING A HEALTH CARE BASED LARGE PROJECT.

This section is based on the summary of the following paper.It is a noted fact that compared to other industries the R&D and capital required in a healthcare, biotech or medical field is expensive. So it is absolutely necessary for project management techniques to be implemented.

[9],[9]

One of the suggestions of the mentioned paper is to have a project management Office

The typical features of a project management office as given by [10]includes

(i)Tracking reporting and information sharing

(ii)Repository of project performance

(iii)Setting Standards

(iv)Project management improvement efforts

(v)Coaching, training internal consulting

(vi)Source of project management talent

(vii)Project cricis response team

[9], [9]

Case study:

A public insurance scheme in the state of Andhra Pradesh India, phase wise implementation. Aarogyasri is a Health Insurance Scheme that IS modeled for the benefit of the poor, to bear the health care costs of families below the poverty line for identified diseases. This scheme was introduced with the guideline that public insurance schemes should be targeted at large scale and fatal illnesses and the benefit in the basic care should be done through free screening and outpatient consultation [11].

1. Aarogyamithras(Facilitator services)
2. Round the clock Call Centre with Toll free help line.
3. Health Camps conducted by network hospitals.
4. Follow up by elaborate field mechanism.
5. End-to-end cashless packages.
6. Services of RAMCO (Rajiv Aarogyasri Medical Coordinator) and AMCCO (Aarogyasri Medical Camp Coordinator) in the network hospitals.
7. CUG (Closed User Group) connectivity to all the field staff, RAMCO and AAMCO.
8. Placement of Aarogyasri kiosk with Network connectivity.
9. Robust IT based solution, capturing patient details right from the reporting to the claim settlement and follow up.
10. Social auditing through feedback letter from the beneficiary and Prajapatham programme.

Aarogyasri to date has screened more than 3.1 million patients in the rural areas and essential drugs were supplied. Further 1.018 million patients were given access to outpatient consultation. The government health system combined with this program Aarogyasri is able to meet the all the health requirements of population in the state. The scheme is managed through effective use of Information Technology based solution which is unique to the scheme in reaching out to the beneficiary. The scheme has many special features to its credit to reach the beneficiary and guide them to use the services without having to pay out of their pocket as given by the government website [11], the following are its salient features

The project was implemented in 3 phases, number of associate hospitals under the scheme 5 per district in the first phase. After successful installation and implementation, phase 2 and phase 3 were implemented.

1.FRAMEWORKS

Distributed Data Mining From Heterogeneous Healthcare Data Repositories: Towards an Intelligent Agent-Based Framework, the mentioned paper [6].

It presents a model with a smart agent based framework for knowledge mining in a distributed healthcare system consisting of many diverse healthcare data centers. Data-centered knowledge mining, especially from many diverse data repositories, is a painstaking process and imposes significant operational constraints on destination users. The independence, interactive and smart agents give an opportunity to create fully blown and accessorized decision support services for healthcare personnel. The use of smart agents for implementation of a distributed Agent based Data Mining Infrastructure that provides a range of healthcare based decision-support systems[6].

Agent-based DM Info-structure (ADMI)

The proposed multi Agent-Based Data Mining Info-Structure (ADMI), “responsible for the generation of data-mediated diagnostic-support and strategic services, takes advantage of a multi-agent architecture which features the amalgamation of various types of intelligent agents, each responsible for an independenttask” [6] .

As given by the reference paper such an agent-federation is designed to service four functional components—

(i)end-user interface;

(ii)remote data access network;

(iii)data mining engine; and

(iv)diagnostic-support and strategic services. [6]

A brief overview of the constituent agents and their functionalities according to the reference paper is summarized as follows:Interface agent: The interface agent is the one who collects user specification for a data mining service.

This collection of user specification is done through a web based interface.

Data Collection agent: The job of a data collection agent is to fetch related data from multiple health care repositories.

Data mining agent: Data mining agent is the one that performs and manages the entire data mining process.Services generation agent: The services generation agent obtains the results from data mining agent and uses it for decision support.

2.CLOUD BASED APPROACH

Cloud computing is the delivery of computing as a service rather than a product, whereby shared resources, software, and information are provided to computers and other devices as a utility (like the electricity grid) over a network (typically the Internet) [12].

Cloud computing provides computation, software, data access, and storage services that do not require end-user knowledge of the physical location and configuration of the system that delivers the services. Parallels to this concept can be drawn with the electricity grid, wherein end-users consume power without needing to understand the component devices or infrastructure required to provide the service [12]. A Cloud Computing Framework for Real-time Rural and Remote Service of Critical Care[12].In a distributed system every node does not have the memory or the computational capacity to maintain all of the systems

3.DATA MINING

Data Mining (the analysis step of the Knowledge Discovery in Databases process, or KDD), a relatively young and interdisciplinary field of computer science, is the process of discovering new patterns from large data sets involving methods from statistics and artificial intelligence but

also database management. In contrast to machine learning, the emphasis lies on the discovery of previously unknown patterns as opposed to generalizing known patterns to new data[12].

4.PRIVACY PRESERVING SHARING OF INFORMATION

Various privacy preserving models in this field have been proposed, Distributed clustering based on sampling local density estimates [13], Secure multi-party computation made simple[14], Privacy-preserving k-means clustering over vertically partitioned data[15], Privacy preserving association rule mining in vertically partitioned data [16], Secure set intersection cardinality with application to association rule mining [17], Privacy Preserving Data Mining, Privacy-preserving data mining [18].

5.HIGH PERFORMANCE COMPUTING FOR DATA MINING

HPC integrates systems administration (including network and security knowledge) and parallel programming into a multidisciplinary field that combines digital electronics, computer architecture, system software, programming languages, algorithms and computational techniques. HPC technologies are the tools and systems used to implement and create high performance computing systems. Recently, HPC systems have shifted from supercomputing to computing clusters and grids. Because of the need of networking in clusters and grids, High Performance Computing Technologies are being promoted by the use of a collapsed network backbone, because the collapsed backbone architecture is simple to troubleshoot and upgrades can be applied to a single router as opposed to multiple ones.

Scalable Data Mining with Log Based Consistency DSM for High Performance Distributed Computing Log Based Consistency Mechanism.

REFERENCES

[1] Privacy Preserving Distributed Learning Clustering Of HealthCare Data Using Cryptography ProtocolsAhmed M. Elmisery, Huaiguo Fu

[2] Interoperability of Medical Device Information andthe Clinical Applications: An HL7 RMIM basedon the ISO/IEEE 11073 DIM.Mustafa Yuksel and Asuman Dogac

[3] Towards a Framework for Health Information Systems Evaluation Maryati Mohd. Yusof, Ray J. Paul, Lampros K. Stergioulas

[4] Research and Implementation of Transmitting and Interchanging Medical Information based on HL7

Xiaoqi LU, Yu GU*, Lidong YANG, Weitao JIA, Lei Wang

[5] Development of Data Authenticity Verification System in Regional Health Information Network

ZHOU Tian-shu, LI Jing-song , ZHANG Xiao-guang, HU Zhen, YU Hai-yan, CHEN Huan

[6] Distributed Data Mining From Heterogeneous Healthcare Data Repositories: Towards an Intelligent Agent-Based FrameworkSyed Zahid Hassan Zaidi Syed Sibte Raza Abidi Selvakumar Manickam

[7] Design and Implementation of Interoperable Medical Information System Based on SOA ZHANG Xiao-guang, LI Jing-song, ZHOU Tian-shu, YANG Yi-bing,CHEN Yun-qi, XUE Wan-guo, ZHAO Jun-ping

[8] Medical Technology Management: From Planning to Application

[10]

[11]

[12] Wikipedia

[13] M. Klusch, et al., "Distributed clustering based on samplinglocal density estimates," presented at the Proceedings of the18th international joint conference on Artificial intelligence, Acapulco, Mexico, 2003.

[14] U. Maurer, "Secure multi-party computation made simple,"Discrete Appl. Math., vol. 154, pp. 370-381, 2006.

[15] J. Vaidya and C. Clifton, "Privacy-preserving k-means clustering over vertically partitioned data," presented at the Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining, Washington, D.C., 2003.

[15] J. Vaidya and C. Clifton, "Privacy preserving association rule mining in vertically partitioned data," presented at the Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining, Edmonton, Alberta, Canada, 2002.}

[16] J. Vaidya and C. Clifton, "Secure set intersection cardinality with application to association rule mining," J. Comput. Secur., vol. 13, pp. 593-622, 2005.

[17] Y. Lindell and B. Pinkas, "Privacy Preserving Data Mining,"presented at the Proceedings of the 20th Annual International Cryptology Conference on Advances in Cryptology, 2000.

[18] R. Agrawal and R. Srikant, "Privacy-preserving data mining," SIGMOD Rec., vol. 29, pp. 439-450, 2000.

All figures in the paper are extracted from the above papers mentioned.