Making The Most Of Data In Order To Provide Accurate Clinical Decision Support Systems For The Use In The Determination Of Heart Disease
Free (open access)
K E Burn-Thornton
This paper provides a description of the results of investigations into the effect of the nature of ECG data, on the classification accuracy of patient state which a Clinical Decision Support System (DSS), based upon the data (knowledgebase), is able to provide. These results show that changes in classification accuracy of 40% can be achieved by changing both the size, and stratification bias, of the knowledgebase. The results suggest that use of a CDSS can be employed to aid an inexperienced physician in the classification of patients with heart disease as long as careful consideration is given to the size, and stratification bias, of the data comprising the knowledge base used to support the DSS. 1 Introduction The current demands placed upon the health service in this country are resulting in not only, a diminishing time being taken to attempt to accurately predict horn the ECGS whether or not heart disease is present but also, the reliance on the less experienced team members to make an accurate decision regarding current patient state. If a decision support system could be provided, with a similar accuracy of patient classification to that shown by ‘experts’ (consultants), not only could ‘expert’ made diagnosis be confined but also those less experienced team members could use the CDSS to support their patient classification.