Atrial fibrillation source area probability mapping using electrogram patterns of multipole catheters

Table 1 Summary of performance of the algorithm for various simulation cases

Simulation type	Catheter type	Successful delineation (%)	Normalized length of trajectory in delineated AF area (mean length)	Difference in PWD and true direction
Rotor source	Lasso	99.98	0.87 ± 0.09 (21.8 mm)	23.76° ± 17.75°
Rotor source	Spiral	99.99	0.80 ± 0.12 (20 mm)	33.80° ± 17.85°
Focal source	Lasso	100	N/A	25.17° ± 15.64°
Focal source	Spiral	100	N/A	34.00° ± 15.77°
Figure of eight	Lasso	97.01	0.66 ± 0.03 (23.1 mm)	23.93° ± 12.41°
Two layer model	Lasso	95.65	0.64 ± 0.26 (12.8 mm)	25.50° ± 19.10°
Conduction block	Lasso	98.57	0.83 ± 0.11 (45.9 mm)	24.54° ± 21.73°
Fibrillatory rotor cluster	Lasso	96.02	0.71 ± 0.29 (20.7 mm)	24.47° ± 20.30°
Rotor source (3D)	Lasso	98.10	0.73 ± 0.10 (10.9 mm)	26.26° ± 19.43°
Focal source (3D)	Lasso	99.41	N/A	25.02° ± 12.04°

The fourth column shows the normalized length of the trajectory of the meandering AF source present inside the highest probability area; for example, a value of 0.87 implies that 87% of the reentry trajectory was present inside the area and the 87% length value is 21.8 mm. Both normalized length and the actual length are averaged over all the algorithm runs. The last column shows the difference between the calculated principal wave direction (PWD) angle and the actual angle in degrees—for reference, the angle between a bipole of a Lasso is about 35° and a spiral has an outer bipole angle of about 56°

ISSN: 1475-925X