GUIDELINES FOR ASSESSING PERFORMANCE OF ST ANALYZERS

This paper proposes principles and methods for assessing performance o f ST analysers and algorithms. We describe an evaluation protocol and performance measures suitable for assessing the accuracy of (1) detect ing episodes of ischaemic ST changes, (2) distinguishing between ischa emic and non-ischaemic ST change episodes, and (3) measuring ST deviat ion and ischaemia duration. There is generally not a one-to-one corres pondence between reference and analyser-annotated ST episodes, nor can non-events be counted. Sensitivity and positive predictivity measures which assess the accuracy of detecting ischaemic ST episodes and tota l ischaemic time are based on the concepts of matching and overlap, re spectively. To address the question of predicting performance in a cli nical environment, we have utilized the bootstrap statistical procedur e, which estimates the mean as well as the standard deviation of the a nalyser's expected performance. We illustrate! the use of the evaluati on protocol and performance measures by a case study in which we prese nt an evaluation of our 2-channel Karhunen-Loeve transform based ST ch ange detection algorithm using the European Society of Cardiology ST-T database.