Healthcare

In healthcare, data mining methods are used along with traditional procedures of mathematical statistics and allow the resolution of a wide range of problems. In this application area all data mining techniques are used: clustering, classification, associative rules, sequential patterns and so on.

The Loginom-based solutions allow the following to be performed:

• Data consolidation. Loginom contains a full-fledged data warehouse which can be used to consolidate all the information associated with the problems under analysis: health records data, results of medical analyses and tests, and output data of diagnostic systems. The powerful integration mechanisms make it possible to receive data from virtually any source, from office applications to specialized medical systems.
• Disease Diagnostics. Loginom can easily be used for creating advisory diagnostic systems that use the accumulated clinical examination data, automatically identify significant characteristics and model complex dependencies between symptoms and diseases. For this purpose, various regression models, decision trees, neural nets, Kohonen maps and others are used.
• Diagnostic test estimation. In medical screening, it is necessary to estimate the effectiveness of diagnostic tests and to compare them with traditional techniques. For this purpose, Loginom uses logistic regression and ROC-analysis. Using these tools, the user can match the optimal thresholds of the diagnostic indicators, and evaluate the sensitivity and specificity of the model.
• Identification of side effects. Data mining techniques such as associative rules and sequential patterns are successfully used for finding the relations between drugs taken and side effects observed. Loginom has convenient tools for resolving such problems.

Loginom is also capable of solving other problems typical of the fields of medicine and biology, namely creation of specialized diagnostic systems, assessment of the effectiveness of new treatment methods, and so on.