Purchase this article with an account.
A.J. Vingrys, J.D. Forte, T.M. Dang, B.V. Bui; Wavelet Analysis of Oscillatory Potentials Following Inhibition of GABAc Receptors . Invest. Ophthalmol. Vis. Sci. 2006;47(13):3094.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We compare three methods of ERG oscillatory potential (OPs) extraction and analysis: filtering with conventional amplitude/implicit time measurement, OP modelling and wavelet analysis of raw data. These methods are applied to consider the effect that inhibition of inner retinal feedback can have on OPs.
Dark–adapted (>12 hrs) full field ERG responses were recorded from anaesthetized (Ketamine:Xylazine, 60:5 mg/kg) adult Long–Evans rats (10–12 wks). Signals (2.7 log cd.s.m–2) were digitised at 4 KHz (0.3 – 1000 Hz) and recorded 45 minutes after intravitreal injection of TPMPA (3 µl, 400 µM, n = 5) to inhibit inner retinal feedback via GABAc mediated mechanisms. The fellow eye was unchanged after treatment and served as control. Waveforms had their Fourier power spectrum defined and either subject to a complex Morlet wavelet analysis (Matlab) to identify frequency characteristics across time or had their P3 components removed to expose post–receptoral contributions. The post–receptoral waveform was then band pass filtered (–6 dB, 50–220 Hz) to give the OP amplitudes and implicit times. OPs were modelled using a GABOR function and these were compared to the relative Wavelet power across time at 10, 20, 80 and 120 Hz.
TPMPA application resulted in a 52± 6% reduction in ERG b–wave amplitude (control, C: 1422 ± 96 vs treated, Tx: 669 ± 80 µV, p<0.01) without a change in implicit time (C: 78 ± 4 vs Tx: 80 ± 5 ms, p=0.75). The photoreceptoral P3 was unaffected by TPMPA (C: –355 ± 45 vs Tx: –378 ± 70 µV, p=0.79). No significant change in OP amplitude (C: 96 ± 13 vs Tx: 88 ± 9 µV, p=0.41) or peak time (C: 29.8 ± 1.1 ms vs Tx: 30.0 ± 1.2 ms, p=0.31) was found with conventional analysis. However, OP modelling found a significant increase in frequency (C: 106.7 ± 5.8 vs Tx: 113.6 ± 5.2 Hz, p=0.005) which was due to a loss of 80 Hz components. Wavelet analysis confirmed a reduction in power in the 80 Hz band (0–90 ms, 21 ± 9%, p=0.04), with no change at 120 Hz (0–70 ms, p=0.41). The increase in OP frequency detected by modelling reflects a selective loss of low frequencies.
Wavelet analysis provides a novel frequency–time OP descriptor. Modelling finds a reduced frequency (∼80 Hz) contribution not evident with conventional methods. Feedback of GABAc receptors modulates low frequency OP components, consistent with the idea that OPs represent multiple mechanisms with distinct frequency characteristics.
This PDF is available to Subscribers Only