Data Mining in Healthcare
- It enables retail sectors to display customer response and helps banking sector to predict customer profitability. It serves many similar sectors such manufacturing, telecom, healthcare, automotive industry, education, and many more.
- The computer keeps a massive amount of patient data with accuracy, and it improves the quality of the whole data management system. Scholars utilizing different clusters, classification, decision trees, neural networks , time series to publish research.
A different approach to mine the data in healthcare:
- Three systems is the way to drive a real-world improvement with any analytics initiative in healthcare. These are the following three systems.
- The analytics system
- The content system
- The deployment system
- Analytics system accepts technology and expertise to accumulate information, comprehend it, and standardize measurements.
- Content system includes standardizing knowledge work.
- It applies evidence-based best practices to care delivery. Strong content system enables organizations to put latest medical conformation practice quickly.
- A strong content system enables organizations to put the latest medical conformation into practice quickly.
- Deployment system involves over new hierarchical structures.
- Deployment system requires a real hierarchical change to drive the adoption of best practices throughout an organization.
Application of Data Mining in Healthcare
- Large amounts of data generated by healthcare transactions are too complex and huge to be processed and analyzed by conventional methods.
- Data mining provides framework and techniques transform these data into useful information for data-driven decision purposes.
- This applications can be used to assess the effectiveness of medical treatments. Data Mining can convey analysis of which course of action demonstrates effective.
- It is used to identify and track chronic illness states and incentive care unit patients, decrease the number of hospital admissions, and supports healthcare management.
Customer Relationship Management
- Customer relationship management is the managing interactions between commercial organizations normally retail sectors and banks, with their customers.
Fraud and Abuse
- It focus on inappropriate or wrong prescriptions and fraud insurance and medical claims.
Results of comparative analysis of various disease in Healthcare:
- The conventional methods of statistical applications are also given and compared.
|S.No||Types of disease||Data mining tool||Technique||Algorithm||Traditional method||Accuracy level % for DM application|
|1.||Tuberculosis||WEKA||Naïve Bayes Classifier||KNN||Probability Statistics||78 Percent|
|2.||Heart Disease||ODND,NCC2||Classification||Naive||Probability||60 Percent|
|3.||Kidney Dialysis||RST||Classification||DecisionMaking||Statistics||76 Percent|
|4.||Diabetes Mellitus||ANN||Classification||C4.5 Algorithm||Neural Network||82 Percent|
|5.||Blood Bank Sector||WEKA||Classification||J48||90 Percent|
|6.||Dengue||SPSS Modeler||C5.0||Statistics||80 Percent|
|7.||Hepatitis C||SNP||Information||Gain||Decision rule||74 Percent|
Advantages of Data Mining in Healthcare
- Healthcare institutions reduce decision-making effort.
- It give the best information support and knowledge to healthcare workers.
- Application of data mining is for biomedical signal processing communicated by internal guidelines and reactions to boost the condition.
Challenges in Healthcare Data Mining
- Components can have a significant effect on the diagnosis, and treatment of a patient. Missing, incorrect, inconsistent data pieces of information saved in various formats from different data sources create a significant obstacle to successful data mining.
- Two major challenges are:
- How to create effective algorithms for differentiating the content of two versions ?
- Demands improvement of effective algorithms and data structures evaluation of knowledge integrity in the data set.
- How to create algorithms for evaluating the impact of specific data modifications on the statistical significance of individual patterns, which is collected with assistance of basic classes of data mining algorithm ?
- Algorithms impact that modifications of data values have on discovered statistical significance of patterns are being created, it is difficult to build up universal measures for all data mining algorithms.