Figure 2.4. Regular sinus rhythm (ventricular bipolar electrogram). It is interpreted as tachycardia intervals (TS), alternating with fibrillatory intervals (FS), except at one interval of 600 ms. This phenomenon is due to sensing of the T wave. This error finally leads to detection of fibrillation (FD).

Early Devices: Detection of Ventricular Fibrillation

The original detection algorithm in the automatic implantable defibrillator (AIDâ„¢) was the probability density function (PDF), conceived to detect sinusoidal rhythms (Intec Systems) [3]. The PDF processed the electrical signal to define the proportion of the cardiac cycle that this signal was deviating from the baseline (Figure 2.5).

The algorithm was based on the observation that the signal during ventricular fibrillation spends the majority of its time away from the isoelectric

Figure 2.5. Probability density function (modified, from Intec) during sinus rhythm (left) and ventricular fibrillation (right).

baseline as compared to the signal during sinus rhythm. CPI shortly proposed a so-called 'turning point morphology' criterion (TPM), derived from the electrogram recorded between the shock leads. The principle was that electrograms with a certain percentage of isoelectric time and a high slew rate would not satisfy the TPM criteria. It was felt that these morphology criteria postponed intervention [4].

Rate-only Detection

The basic goal of the ICD is to detect and subsequently terminate life-threatening ventricular tachyarrhythmias. The most fundamental criterion for the detection of ventricular tachyarrhythmias is based on rate. This is measured by assessing the duration of the cardiac cycle length (time divided by rate) on a beat-to-beat basis. Each detected ventricular interval is compared with the programmed detection zones. This basic detection algorithm measures and counts ventricular intervals to detect a tachyarrhythmia if it fulfils the criteria of ventricular rate over a certain, predefined duration. With correct sensing, this method ensures 100% sensitivity of ventricular tachyarrhythmias with rates above the programmed detection rate. However, rate-only detection has a poor specificity in arrhythmia discrimination. The reported incidence of inappropriate therapy ranged between 16 and 41% [5-8]. This incidence may have been underestimated due to the lack of electrogram storage. The definition of 'appropriate' therapy relied on concomitant ECG monitoring. The fact that atrial tachyarrhythmias contributed to the incidence of inappropriate therapy was established.

With tiered-therapy devices (providing antitachycardia pacing for slower tachycardias), the inappropriate detection of atrial tachyarrhythmias became a greater problem [9]. This is due to the increased probability of rate overlap between the target ventricular tachyarrhythmias and atrial tachyarrhythmias, as lower detection zones can be programmed.

Detection Zones

In modern ICDs, the range of ventricular rates is divided into a bradyarrhythmia, a normal, and up to three tachyarrhythmia detection zones (Figure 2.6). In all ICDs, the highest detection zone is called the 'fibrillation' zone. In order for a tachyarrhythmia to be detected and assigned to a given detection zone, it must exhibit a certain number of intervals (duration). For this duration criterion different methods of counting are used (consecutive interval counting, probabilistic counting, or a combination of these).

Probabilistic counting algorithms are used for the detection of ventricular fibrillation. This algorithm requires a defined proportion of ventricular intervals within a sliding window to be shorter than the programmed detection interval for fibrillation (X of Y counter). This method reduces the chance of underdetection of ventricular fibrillation due to the irregularity of ventricular intervals and the continuously changing amplitude of the signal. In Medtronic ICDs, for example, 75% of consecutive intervals must be within the fibrillation detection zone. In the Guidant algorithm, a fixed window width of 10 intervals is used. Initially, 8 of 10 intervals are required within the fibrillation detection zone for rate detection to be met, and at least 6 of 10 intervals must remain in the detection zone for a programmed detection time.

The counting algorithm for ventricular tachycardia may be either probabilistic or consecutive in design. The consecutive-interval algorithm diminishes the risk of inappropriate therapy for atrial fibrillation, without compromising the sensitivity for detection of ventricular tachycardias. In consecutive-interval algorithms, the counter increments every time when an interval is measured within the detection zone. An interval outside the detection zone will reset the counter to zero. For patients with both slow and fast ventricular tachycardias, the programming of two tachycardia detection zones allow zone-specific detection and therapies. The programming of slow tachycardia detection zones increases the risk of inappropriate therapies for atrial tachyarrhythmias [9,10].

Fibrillation zone

Was this article helpful?

0 0

Post a comment