Abstract:
There is a huge amount of ‘knowledge-enriched data’ in hospitals, which needs to be
processed in order to extract useful information from it. The knowledge-enriched data is very
useful in making valuable medical decisions. However, there is a lack of effective analysis
framework to discover hidden relationships in data. The objective of this research is to propose a framework for cardiac data mining that mines the historical unstructured data of heart
patients and extract significant features and patterns which will not only enable doctors to
predict heart attack but also provide in-depth insight to write better prescription in future.
This work is based on a large amount of unstructured data in the form of patients medical
reports collected from a renowned cardiac hospital in Pakistan. Firstly data preparation is
done in which the unstructured (textual) data of heart patients is converted to structured (tabular) form and then pre-processed to make it suitable to apply different data mining techniques. After data preprocessing, data mining techniques are used in which clustering, correlationand association rule mining techniquesare applied onthe dataset. The output from this
exercise takes the form of trends, patterns and rules which can then be used for heart condition prediction besides helping medical practitioners in making intelligent verdicts. Finally,
performance evaluation of the selected k-Means algorithm is performed with other clustering
algorithms on the basis ofsome internal evaluation indexes.Further the generated rules are
evaluated using statistical measures such as support, confidence, lift, completeness, interestingness and comprehensibility to extract significant rules only
