Identification of low-voltage areas during atrial fibrillation: a unipolar, bipolar and omnipolar perspective
M S Van Schie, C M W Kuis, A I Freriks, N L M De Kruijf, P Knops, V Yildirim, Y J H J Taverne, N M S De GrootAbstract
Background
Voltage maps are often used to identify target sites for substrate ablation during atrial fibrillation (AF). In daily clinic, bipolar electrograms (EGM) are frequently used, although their directional dependency can result in misleading potential voltages, particularly during AF. Usage of omnipolar technology overcomes these limitations by providing a direction independent omnipolar EGM. However, the relationship between the derived omnipolar and the corresponding bipolar potential voltages during AF remains unknown. In addition, it is unknown whether unipolar EGMs provide complementary or contradictory information on the AF voltage maps.
Purpose
To examine (dis)similarities in unipolar, bipolar and omnipolar voltage mapping during AF.
Methods
Intra-operative high-resolution epicardial mapping (1760±263 10-second EGMs per patient) with a 128- or 192-electrode array (interelectrode distance 2 mm) was performed in 8 patients undergoing coronary artery bypass grafting. Bipolar EGMs were created from adjacent unipolar EGMs. Omnipolar EGMs were created from perpendicular bipolar EGM pairs. The maximum voltage of 3 unipolar, corresponding 2 bipolar and omnipolar potential (clique) were compared. Low-voltage potentials (LVP) were defined as unipolar voltage <1 mV and bipolar and omnipolar voltage <0.5 mV.
Results
In total, the data set consisted of 1,426,917 cliques (178,365±38,377 per patient). Unipolar voltages were larger than both bipolar and omnipolar voltages (2.4 [1.2-4.6] vs 1.1 [0.4-2.9] & 1.2 [0.4-3.2] mV, p<0.0001 both). The maximum bipolar voltage was on average 1.5 (ranging from 1.0-252.0) times larger than the corresponding bipolar voltage in perpendicular direction. Unipolar voltages had an inverse quadratic relationship with a linear component with bipolar and omnipolar voltages (R2=0.83, p<0.0001 and R2=0.82, p<0.0001 respectively). A linear correlation was observed between bipolar and omnipolar voltages (R2=0.98, p<0.0001). Using bipolar clique voltages, 31.5% LVPs were found, while this number increased to even 37.9% when taking the corresponding perpendicular bipolar voltages. On the other hand, the percentage of LVPs was considerably lower using omnipolar and unipolar clique voltages (28.3% and 19.3% respectively); 15.7% of all cliques contained both unipolar and omnipolar LVPs.
Conclusions
Considerable directional differences in bipolar voltages exist during AF. Direction independent omnipolar voltages are therefore preferred to prevent overestimation of bipolar low-voltage areas. However, a combination of unipolar and omnipolar voltages could be even more precise in identifying true low-voltage areas during AF.Examples of AF electrogramsRelationships between various voltages