Advanced detection of ST segment episodes in 24-hour ambulatory ECG data by automated tracking of transient ST segment reference level

Abstract

Using the Long-Term ST Database, we developed and evaluated an advanced algorithm for automated detection of transient ST segment episodes in "real-world" 24-hour ambulatory data. To successfully detect transient ST change episodes, the algorithm automatically tracks the time-varying ST segment reference level due to clinically not important non-ischemic causes and subtracts it from the ST segment level. Evaluating of the algorithm using reference annotations of the protocol B of the database yielded gross ST episode detection sensitivity and positive predictivity of approximately 75%.