Contactless recordings of retinal activity using optically pumped magnetometers

Optically pumped magnetometers (OPMs) have been adopted for the recording of brain activity. Without the need to be cooled to cryogenic temperatures, an array of these sensors can be placed more flexibly, which allows for the recording of neuronal structures other than neocortex. Here we use eight OPM sensors to record human retinal activity following flash stimulation. We compare this magnetoretinographic (MRG) activity to the simultaneously recorded electroretinogram of the eight participants. The MRG shows the familiar flash-evoked potentials (a-wave and b-wave) and shares a highly significant amount of information with the electroretinogram recording (both in a simultaneous and separate recording). We conclude that OPM sensors have the potential to become a contactless alternative to fiber electrodes for the recording of retinal activity. Such a contactless solution can benefit both clinical and neuroscientific settings.

1 Introduction the data stream was recorded with both the OPM and MEG system for offline synchronization 72 of the recordings. 73 To actively cancel the magnetic gradients produced by the magnetic parts of the Elekta MEG 74 system situated in the shielded room, we used a Helmholtz coil around the magnetic parts of 75 the MEG dewar. The gradients at the location of the measurement were nulled before the 76 measurement started, using a fluxgate as feedback for manual adjustments of the currents sent 77 to the Helmholtz coil. Immediately before the start of the recording, the OPM sensors measured 236 cd m −2 on the screen, and the room was dark during the presentation of the stimuli. 88 One set of stimuli consisted of 400 trials, and every participant viewed two sets. For the first 89 set, the OPM sensors and the DTL fiber electrode were used for recording, for the second one, 90 the fiber electrode was removed from the participant's eye and only the OPM sensors were in 91 place. This was done to have both simultaneously recorded data from ERG and MRG channels 92 as well as a recording from the OPM sensors alone to ensure that any potential retinal activity 93 measured by the OPMs was not introduced by the ERG electrode.  (Oostenveld et al., 2011)  The data were high-pass filtered at 1 Hz (onepass-zerophase, hamming-windowed FIR filter 105 with order 1650) and low-pass filtered at 45 Hz (onepass-zerophase, hamming-windowed FIR 106 filter with order 294). The data were then baseline corrected using the time window of −100 ms 107 to 0 ms before flash-onset as baseline.

108
Mutual information To quantify the shared information between the ERG and MRG, mu-109 tual information was computed within subjects between the ERG electrode and an OPM chan-110 nel. The OPM channel was selected per data set and participant by searching for the highest 111 b-wave amplitude across channels (search window for maximum: 0 ms to 100 ms). The selection 112 of one "best" channel per participant and data set was motivated by the fact that the placement 113 of the OPM array could slightly vary per recording session and participant. Next, we computed 114 the mutual information of the MRG channel with the ERG electrode for the time window of 115 −50 ms to 150 ms per subject and data set. The computation of the mutual information was 116 based on Cohen (2014). The optimal number of bins for the computation of the mutual infor-117 mation was estimated using the Freedman-Diaconis rule. On average, 8.94 bins were estimated 118 to represent the data optimally (median 9.0, SD 2.14, range 6 − 13), thus the number of bins 119 was set to 9 for all data sets and participants. The mutual information is reported in bits. A 120 permutation test was performed to determine whether the two time series shared information.

121
To this end, one of the two time series was shifted relative to the other by a random amount.

122
This procedure was repeated 5000 times to create a null distribution, which was then used to 123 estimate the p-values.

124
Estimation of signal and noise properties The noise spectrum of the data was computed 125 on the continuous data of the simultaneous data set before any other processing of the data 126 was done. The power spectral density was estimated for a range of 1 Hz to 200 Hz in 1 Hz steps 127 using the Fast Fourier Transform with a Hamming window. The signal-to-noise ratio (SNR) 128 was estimated on the epoched data after line noise removal and artifact rejection. The SNR 129 estimate was computed using the root-mean-square of a baseline (−100 ms to −30 ms before 130 flash onset) and active period (30 ms to 100 ms after flash onset) and then converted to decibels.

132
We recorded retinal activity with 8 OPM sensors during the presentation of two sets of short 133 flash stimuli. Figure 1 shows the recorded activity for the OPM sensors, averaged across partici-  A depicts the simultaneous recording, when the ERG electrode was also present. B shows the data from the second recording from the same participants after the ERG electrode was removed. Figure 2A shows the single subject traces from the simultaneous recordings, averaged across 143 all trials and channels. Also here, the a-wave and b-wave can be identified in all participants 144 and gets even more pronounced when looking at the trial-averages of the best channel (defined 145 as the channel with the highest b-wave amplitude) in Figure 2B. The ERG averages ( Fig. 2) 146 show a high resemblance to the single subject averages, but also to the data from both the 147 simultaneous and separate recording shown in Figure 1.
148 Figure 3A shows  S1). 168 We also computed the mutual information between the ERG and each OPM sensor for  Fig. 2 B) and recording (the solid line corresponds to the simultaneous recording with the ERG, and the dashed line depicts the data from the separate OPM recording). B The bar graph shows the mutual information between the best OPM channel and the ERG per participant. The light green bars represent the simultaneous MRG-ERG recording, and the dark green bars the separate, MRG-only recording.  Figure 4: Sensitivity of the OPM array. Shown is the mutual information per channel with the ERG activity of the simultaneous recording, averaged across subjects. The color corresponds to the mutual information, the size of the circle is inversely proportional to the standard deviation across participants (such that smaller circles correspond to a less certain estimate of the mutual information).
shows the precision of this estimate: the size is inversely proportional to the standard deviation  Lastly, we looked at the signal and noise properties of both the MRG and ERG signals by 180 computing the noise spectrum as well as SNR. Figure 5A compares the noise spectrum of 181 the ERG (orange) and MRG (blue) across the continuous data of the simultaneous recording.
182 Figure 5B depicts the signal to noise ratio of the single trials.   196 In this paper, we recorded retinal activity with OPM sensors for the first time. We show that 197 the recorded MRG activity exhibits both the a-wave and b-wave (Fig. 2), which are the typ-198 ical retinal potentials following visual flash stimulation. We also demonstrate that the MRG 199 activity closely matches the retinal activity as recorded with ERG fiber electrodes in the same 200 participants (Fig. 3). Furthermore, we describe the distribution of the magnetic field of retinal 201 activity over our recording array (Fig. 4) and defined the signal and noise characteristics of the 202 ERG and MRG data (Fig. 5).

207
we show that with OPMs, we are not restricted to (sub-)cortical brain activity, but can also 208 easily record the magnetic activity of the retina. Our OPM recordings share a highly signifi-209 cant amount of information with the ERG recording and both the a-wave and the b-wave are 210 clearly visible in the evoked activity. Compared to the seminal paper by Katila et al. (1981), 211 who reported an a-wave to b-wave amplitude of approximately 0.1 pT, our measurements show 212 an increase of roughly 6-fold (cf. Fig. 3A and Fig. 14 in Katila et al., 1981).
213 Surprisingly, the shared information between the MRG and ERG was slightly higher for the 214 MRG data from the separate recording compared to the MRG from the simultaneous recording.

215
This effect was not driven by the difference in the amount of trials that was removed due to 216 artifacts, which was higher in the simultaneous recording. Possibly, this result can be explained 217 by an on average slightly more favourable position of the OPM sensors for recording retinal 218 activity in the separate recording or by a slightly poorer SNR in the simultaneous recording 219 through artifacts induced by tiny motions of the ERG cable. Also note, that the Bayes Factor 220 of this result is not very high, and is only classified as "anecdotal evidence" (Jeffreys, 1961;221 Lee and Wagenmaker, 2014). Ultimately, we can still conclude that the retinal activity in the 222 OPM sensors is not simply induced by the ERG electrode, but genuine recorded activity. cheek of the participants. This matches the early report of MRG recordings by Katila and 228 colleagues (1981), who also found the area below the eye to be the most sensitive to retinal 229 activity in their recordings with a SQUID magnetometer (cf. Fig. 15 in their paper).

230
We placed the grid with the sensors in reference to the participants' eyes, thus, the exact place-231 ment naturally may have varied across participants. With more exact placement, a holder for 232 the OPMs that follows the anatomy more closely, and co-registration of the OPMs to facial 233 landmarks, the signals should be even more comparable across participants. Then, it would 234 even be possible to look at the topographical distribution of the signal, e.g. in combination 235 with multifocal ERG stimulation (Derafshi et al., 2017).

237
While the OPM recordings have an overall lower SNR, they show less artifacts. We report 238 a significant difference between the MRG and the ERG recordings for the number of trials that 239 had to be rejected due to eye blinks or other artifacts. The ERG electrode has direct contact to 240 the lower eye lid, and thus gets inevitably moved during blinks or eye movements, causing large 241 artifacts or even signal saturation. Since the OPM array did not have any physical contact 242 with the participant, blinks or eye movements do not cause the same disturbances in the signal.

243
This is a clear advantage of the OPM-MRG over the fiber electrode ERG.

244
The SNR of the OPM-MRG recordings could be improved by further decreasing the noise in 245 the shielded room, the background fields in our experiment were in the range of up to 15 nT.

246
A placement of the OPM sensors around the eye that is better adjusted to the head's shape record the oscillatory potential, which is centered around 120 Hz (Frishman, 2013 Figure S1: Mutual information between the ERG and MRG recordings. This figure shows the mutual information between the best OPM channel and the ERG per participant. The dark green bars represent the simultaneous MRG-ERG recording, the mid green bars the separate, MRG-only recording. The light green bars represent the MRG-only recording after randomly subsampling the data to include the same number of trials as the ERG data from the simultaneous data set.