A Case Study of Data Mining Algorithms in Healthcare

Utilizing Data Mining Algorithms for Efficient Healthcare Decision Making

by Nisha Rani*, Dr. Y. P. Singh,

- Published in Journal of Advances and Scholarly Researches in Allied Education, E-ISSN: 2230-7540

Volume 14, Issue No. 2, Jan 2018, Pages 680 - 684 (5)

Published by: Ignited Minds Journals


ABSTRACT

Data Mining is defined as the procedure of extracting information from huge sets of data or mining knowledge from data. Data mining helps the healthcare systems to use data more efficiently and effectively. This paper reviews on the implementation of data mining techniques in health sector. Data mining provides the methodology and technology to transform huge amount of data into useful information for decision making. This paper explores data mining applications in healthcare. Information digging is a procedure for finding valuable data from vast databases. It draws thoughts from number of controls, for example, measurements, AI and database frameworks. Diminishing expense of electronic stockpiling media has made it financially possible for human services associations to keep up vast therapeutic databases. Examination of huge medicinal databases can help in getting affiliation rules, showing connection between methodology performed on patients and the detailed conclusion. Present day drug produce enormous measure of patient medicinal services information. These information records are put away electronically by therapeutic network in databases. The data contained in medicinal information records is fascinating and valuable for patient consideration. Information mining strategies are calculations that are utilized for structure models and for discovering designs in information. In medicinal information mining building exact classifier demonstrate for anticipating genuine human maladies is critical. Classifier models may help a doctor to precisely analyze a malady.

KEYWORD

data mining, healthcare systems, implementation, methodology, technology, data, decision making, applications, information digging, databases, association rules, patient healthcare data, algorithms, classifier model, diagnose

INTRODUCTION

Knowledge discovery in databases is well-defined process consisting of several distinct steps. Data mining is the core step, which results in the discovery of hidden but useful knowledge from massive databases. A formal definition of Knowledge discovery in databases is given as follows: ―Data mining is the non-trivial extraction of implicit previously unknown and potentially useful information about data‖. Data mining technology provides a user oriented approach to novel and hidden patterns in the data. The discovered knowledge can be used by the healthcare administrators to improve the quality of service. The discovered knowledge can also be used by the medical practitioners to reduce the number of adverse drug effect, to suggest less expensive therapeutically equivalent alternatives. Anticipating patient‘s future behavior on the given history is one of the important applications of data mining techniques that can be used in health care management.

INTER-RESIDENT ASSESSMENT INSTRUMENT (INTER-RAI)

The Inter-RAI is an extensive institutionalized instrument for assessing the requirements, qualities and inclinations of mental patients in institutional settings. Between RAI go for patients with intense consideration and long haul needs. Between RAI comprises of an accumulation of patient evaluation instruments, which are utilized to assemble data, for example, patient's qualities and needs, and are likewise used to create singular consideration plans for various patients. These evaluations can be refreshed by the patients' health which ought to improve the consideration that is given to the patient. The Inter-RAI is essentially an organized thought of how to deliver a very much characterized way to deal with recognize the issue as for treating a patient who requires long haul care. There are in excess of eight unique sorts of Inter-RAI appraisal instruments. These arrangement of evaluations are redone as per the patients necessities, hence not every one of the patients will have a similar appraisal structure, which implies a patient with intense consideration needs concerning seniority offices will have diverse

12

connected for a specific evaluation. In Inter-RAI there are various structures that are required for determination comparing to certain healthcare issues, for example, with some intense consideration or finding of patients with psychological well-being. The Inter-RAI gathering of instruments is additionally a sort of least dataal index instruments. This can be considered as the base number of inquiries that are required to make an appropriate analysis of a patient as for a specific intense issue. All very much characterized issue recognizable proof procedure pursues comparable strides as referenced underneath where RAP is the occupant evaluation principle.

Figure 1: Assessment configurations for the Inter-RAI system

The final product of executing these structures is, improved inhabitant care and better personal satisfaction because of the intensive analysis of the patient with the assistance of the Inter-RAI frames. Expanding consideration gave to every inhabitant should result in the patient reacting better to treatment. Clinical staff will have a clearer picture having every one of the documentations of the patient close by and in this manner creating successful correspondence between staff individuals and individual inhabitants. The documentation of the Inter-RAI is clear and there will be just a single response to each question. With appropriate documentation there ought to be less administrative blunders and, in the meantime, instructing new staff individuals will be simpler.

HEALTH ANALYTICS

Healthcare organization can improve their efficiency and effectiveness in the delivery of their duties as it relate to data extraction from patterns in other to making accurate decision by the use of information technology, computer based tools, mathematical computation, statistical tools etc. in health analytics decision making and problem solving is very vital for proper care of the patients. It enables the practitioners and health workers to make policy, improve working condition with d use of the technological base support tools. Health analytics stages start with collection, preprocessing, transformation of health data. develop and implement in healthcare. Descriptive analytics uses graphical representation for better understands by practitioners in the field. Descriptive analytics gives overall details of number of patients treated, revenue generated, what are the symptoms of the patients, diseases diagnosed, how are they treated and managed to improve their condition. It gives a summary of the historical data for generate meaningful information. b. Predictive analytics this gives a focus to the use of information. Predictive analytics which is used to identify future probabilities and trends to predict future occurrences. It ask a question of what could happen in the future. It gather information from historical background, learn patterns from the dataset to be able to predict the future by extract knowledge. Volumes of complicated data available in healthcare allow predictive analytics of the techniques used in data mining. Health professional could ask, what drugs to use for the treatment? Who could be affecting by this diseases next? To predict result of a patient and allocate resources appropriately. c. Prescriptive analytics are applied when there are several options in the health problems or choice to deliver the best prescriptive analytics. Prescriptive analytics ask, how do we respond to those potential future events? This has been used in healthcare for treatment and drug prescription. Several drugs might be prescribed by weighing the pros and cons. It can be used to determine most accurate solution for a given problem in a dataset features. d. Discovery analytics make use of the discovered knowledge to come up with new invention and innovation in the field of data mining healthcare. For examples, from previously known drugs to discover new drugs. This can help to discover new treatment, new diseases, and new medication from the known discovery. Data discovery is used in some healthcare organization to optimize processes that in a given dataset features. Many application of data mining involves all the four health analytics.

Nisha Rani1* Dr. Y. P. Singh2

Each quality in the MDS structure can be considered as an inquiry that required to be offered an explanation to evaluate a patient's needs. A few times the data that is acquired for a specific characteristic won't be adequate for appropriate total appraisal, in this way we have to give more data respect to this specific quality. In this way RAPS can be utilized to furnish singular consideration to every patient as for social, medical and mental issues. • Utilization Guidelines This can be considered as the documentation of the RAI system. Therefore there will be no misconception with respect to qualities and processing that should be given to newcomers for finishing the RAI-MDS shapes. This is essential as this will help counteract misconception or deception of qualities amid the structure filling systems. There are numerous types of RAI that have been grouped for various divisions of health services. These are a lot of structures that will help appropriate evaluation of a patient. A portion of the distinctive kinds of evaluation instruments are as referenced underneath • RAI 2.0 utilized for appraisal in constant consideration/nursing home • RAI-HC utilized in home consideration • RAI-MH utilized in analysis of emotional well-being • RAI-AC for Acute consideration • RAI-PAC Post-Acute Care-Rehabilitation The upside of the RAI system is that they are incorporated with each other. There are various applications for the RAI systems. RAI/MDS data is for the most part utilized for consideration arranging, deciding quality pointers, result estimation, case-blend based financing and deciding qualification for administrations.

HEALTH DATA SOURCE AND TRANSFORMATION

Health data sources: Health data can be collected from both hospitals and healthcare practitioners. Hospital data include patient‘s data, diagnosis, drug prescription and treatment data while healthcare practitioners‘ data are data collected from government Organization (WHO) and other healthcare organizations. The most reliable healthcare data comes from governmental sources or healthcare professional organizations. The data comes from several sources of varying quality, inconsistent and incorrect that are of immense volume. The volume of the data needs to be organized, structured and processed to get meaningful result. There are five most useful sources of data which are clinical data, claims data, patient-generated data and pharmaceutical data.‘ Health data transformation: data transformation is the process of changing data to information, usually from one format to another. The format from the source data are changed to the format of the desired information. Data needs to be transforming before it can make more sense. Data in a raw state is not meaningful and useful until it‘s transformed. Different approach can be employ to transform data such as (i) service-oriented architectural (ii) Data warehousing. Data are made available in different form which is executed incrementally. Commonly used transformational languages are Perl, AWK, TEXT, XSLT and template languages and processors. DATA MINING APPLICATIONS IN HEALTHCARE SECTOR: Healthcare industry today generates large amounts of complex data about patients, hospital resources, disease diagnosis, electronic patient records, medical devices etc. Larger amounts of data are a key resource to be processed and analyzed for knowledge extraction that enables support for cost-savings and decision making. Data mining applications in healthcare can be grouped as the evaluation into broad categories (Prasanna, 2011).

Fig 2: Data Mining Architecture

12

mining can deliver an analysis of which course of action proves effective by comparing and contrasting causes, symptoms, and courses of treatments. 2. Healthcare management: Data mining applications can be developed to better identify and track chronic disease states and high-risk patients, design appropriate interventions, and reduce the number of hospital admissions and claims to aid healthcare management. Data mining used to analyze massive volumes of data and statistics to search for patterns that might indicate an attack by bio-terrorists. 3. Customer relationship management: Customer relationship management is a core approach to managing interactions between commercial organizations-typically banks and retailers-and their customers, it is no less important in a healthcare context. Customer interactions may occur through call centers, physicians‘ offices, billing departments, inpatient settings, and ambulatory care settings. 4. Fraud and abuse: Detect fraud and abuses establish norms and then identify unusual or abnormal patterns of claims by physicians, clinics, or others attempt in data mining applications. Data mining applications fraud and abuse applications can highlight inappropriate prescriptions or referrals and fraudulent insurance and medical claims. 5. Medical Device Industry: Healthcare system‘s one important point is medical device. For best communication work this one is mostly used. Mobile communications and low-cost of wireless bio-sensors have paved the way for development of mobile healthcare applications that supply a convenient, safe and constant way of monitoring of vital signs of patients (Haghighi, et al 2005). Ubiquitous Data Stream Mining (UDM) techniques such as light weight, one-pass data stream mining algorithms can perform real-time analysis on-board small/mobile devices while considering available resources such as battery charge and available memory. 6. Pharmaceutical Industry: The technology is being used to help the pharmaceutical firms manage their inventories and to develop new product and services. A deep understanding of the knowledge hidden in the Pharma data is vital to a firm‘s competitive position and organizational decision-making. data. Application of data mining to data stored in a hospital information system in which temporal behavior of global hospital activities is visualized. Three layers of hospital management:

  • Services for hospital management
  • Services for medical staff
  • Services for patients 2.8 System Biology: Biological databases contain a wide variety of data types, often with rich relational structure. Consequently multi-relational data mining techniques are frequently applied to biological data (David et al., 2008). Systems biology is at least as demanding as, and perhaps more demanding than, the genomic challenge that has fired international science and gained public attention.

CONCLUSION:

Data mining algorithms are providing huge success in the field of healthcare particularly for the diagnosis of diseases. Lot of work has been done and lots more to be accomplished in future. Every algorithm that has been worked upon has shown their individual capabilities in predicting diseases and there is lot of variation depending upon the attributes. The result of our analysis and after studying the work of various authors shows that the algorithms namely naïve bayes, SVM, decision tree, KNN and neural network are the most commonly used algorithms for diagnosis of diseases and thus hold grave importance in healthcare. Data mining tools used in healthcare to predict future outcome from information generated that assist health organizations to make decision. The descriptive, predictive, prescriptive and discovery analytics were introduced. Sources and transformation of health data were discussed. Related works of previous research were reviewed and specific application areas of data mining in healthcare were mentioned.

REFERENCES

1. H. Mamiya, K. Schwartzman, A. Verma, C. Jauvin, M. Behr, and D. Buckeridge (2015). ―Towards probabilistic decision support in public health practice: Predicting recent transmission of tuberculosis from patient attributes,‖ J. Biomed. Inform, vol. 53, pp. 237–242 2. V. L. S. Thompson, S. Lander, S. Xu, and C. Shyu(2014). ―Identifying key variables in African American adherence to colorectal

Nisha Rani1* Dr. Y. P. Singh2

mining,‖ pp. 1–10 3. S. Bandyopadhyay, J. Wolfson, D. M. Vock, G. Vazquez-Benitez, G. Adomavicius, M. Elidrisi, P. E. Johnson, and P. J. O‘Connor (2014). Data mining for censored time-to-event data: A Bayesian network model for predicting cardiovascular risk from electronic health record data. 4. B. Zheng, S. W. Yoon, and S. S. Lam(2014). ―Breast cancer diagnosis based on feature extraction using a hybrid of K-means and support vector machine algorithms,‖ Expert Syst. Appl., vol. 41, no. 4 PART 1, pp. 1476–1482. 5. S. Kang, P. Kang, T. Ko, S. Cho, S. Rhee, and K.-S. Yu (2015) ―An efficient and effective ensemble of support vector machines for anti-diabetic drug failure prediction,‖ Expert Syst. Appl., vol. 42, no. 9, pp. 4265–4273 6. S. H. Liao, P. H. Chu, and P. Y. Hsiao (2012). ―Data mining techniques and applications - A decade review from 2000 to 2011,‖ Expert Syst. Appl., vol. 39, no. 12, pp. 11303–11311 7. R. Veloso, F. Portela, M. F. Santos, Á. Silva, F. Rua, A. Abelha, and J. Machado (2014). ―A Clustering Approach for Predicting Readmissions in Intensive Medicine,‖ Procedia Technol., vol. 16, pp. 1307–1316. 8. C. T. Su, P. C. Wang, Y. C. Chen, and L. F. Chen (2012). ―Data mining techniques for assisting the diagnosis of pressure ulcer development in surgical patients,‖ J. Med. Syst., vol. 36, no. 4, pp. 2387–2399 9. H. M. Zolbanin, D. Delen, and A. Hassan Zadeh (2015). ―Predicting overall survivability in comorbidity of cancers: A data mining approach,‖ Decis. Support Syst., vol. 74, pp. 150–161 10. P. J. García-Laencina, P. H. Abreu, M. H. Abreu, and N. Afonoso (2015). ―Missing data imputation on the 5-year survival prediction of breast cancer patients with unknown discrete values,‖ Comput. Biol. Med., vol. 59, pp. 125–133 11. H. Mamiya, K. Schwartzman, A. Verma, C. Jauvin, M. Behr, and D. Buckeridge (2015). ―Towards probabilistic decision support in public health practice: Predicting recent transmission of tuberculosis from patient

237–242

12. V. L. S. Thompson, S. Lander, S. Xu, and C. Shyu(2014), ―Identifying key variables in African American adherence to colorectal cancer screening: the application of data mining,‖ pp. 1–10

Corresponding Author Nisha Rani*

Research Scholar of OPJS University, Churu, Rajasthan