Effects of electrode misplacement on the reconstruction of the 12-lead ECG

In this study we investigate a limited lead system, that reconstructs 12-lead ECGs from leads I, II, V2 and V5, to assess how slight misplacement of recording electrodes impacts on reconstruction accuracy. The study population consisted of 117 lead body surface potential maps (BSPMs) recorded from 559 subjects (approximately one third normal, one third MI and one third LVH). The BSPMs were interpolated to increase the number of recording sites in the vicinity of V2 and V5. For QRS segments the median RMS error across all reconstructed leads was 220.4 ¿V, 171.4 ¿V, and 277.8 ¿V when V2 and V5 were simultaneously moved -50 mm vertically, 0 mm, and +50 mm vertically respectively. For STT segments these values were 66.8 ¿V, 54.3 ¿V and 76.9 ¿V respectively. We observed that during the QRS segment the most accurate reconstruction was at -15 mm (RMS error: 154.4 ¿V). During the STT segment the reconstruction error was at its minimum at -20 mm (RMS error: 48.5 ¿V). A similar increase in performance for STT reconstruction was observed at 15 mm (RMS error: 48.6 ¿V). The median values taken across all leads masked the fact that electrode misplacement affected different reconstructed leads in different ways.