4.2 Response time reflects the mechanism of ephaptic coupling
Response times in our study are measured as the latency between the onset of stimulation and the keypress from participants reporting seeing a flash. Our findings reveal that multiple factors influence response time, including stimulation polarity, amplitude modulation, and intensity. Participants exhibited slower responses in cathodal otDCS compared to tACS, in 2 Hz amplitude-modulated at 18 Hz compared to sinusoidal at 18 Hz, and in threshold compared to suprathreshold stimulations.
Among these findings, the effects of both intensity and AM can be explained by the amount of delivered energy due to waveform shape and amplitude. With the same peak-to-peak amplitude, AM stimulation could deliver approximately half the energy of sinusoidal stimulation. In our experiment, the threshold amplitude for AM stimulation is about 1.3 times that of sinusoidal stimulation, with the delivered energy being approximately 65% of the threshold sinusoidal stimulation.
While the relationship between stimulation intensity and response time is commonly observed in psychophysics studies, the underlying neural mechanisms are rarely explained. Expanding on the global resonance theory, our results can be elucidated by the occurrence of ephaptic couplings in phosphene perception. Unlike typical synaptic transmission, ephaptic coupling transfers information between adjacent neurons through extracellular ion exchange or potential oscillation. It is thought to influence the synchronization of neural spiking (Anastassiou et al., 2011) and could be potent enough to elicit the firing of adjacent pyramidal cells with a single action potential (Han et al., 2018).
Computational models propose that ephaptic coupling within neural bundles may explain selective behavior (Chawla & Morgera, 2014), responses to peripheral nerve stimulation (Capllonch-Juan & Sepulveda, 2020), and the propagation of epilepsy (Shivacharan et al., 2021). A recent simulation study (Schmidt et al., 2021) demonstrated how ephaptic couplings could contribute to response time. The model showed that stronger stimulation triggers more neural spiking. While each spike could induce ephaptic coupling within the neural bundle, collections of ephaptic couplings become strong enough to elicit spiking volleys, thereby increasing the velocity of neural transmission that leads to the acceleration of response time. According to the model, not only are more neural spiking elicited by stronger stimuli, but the stimulation effect is also prolonged in the neural system.
This ephaptic coupling hypothesis could also explain the slower response time in cathodal otDCS. It is possible that cathodal direct current creates a hyper-polarized environment that makes it difficult to maintain the spiking volley. Therefore, our response time results can be explained by ephaptic coupling, which is considered the neural mechanism of the global resonance theory.
Polarity effect may imply the selective effect of stimulation
One of the primary goals of our study is to examine the polarity effect on phosphene perception. Compared to tACS, anodal otDCS induced brighter phosphene, but no significant difference was found between cathodal otDCS and tACS. However, cathodal otDCS did lead to longer response times. Interestingly, none of the otDCS conditions influenced the phosphene threshold, which is typically considered an index of neural excitability. This result indicates that while the occurrence of phosphenes is determined by electromagnetic oscillation, the perceptual quality can be influenced by cortical excitability.
TES has been observed to influence visual perception in various ways. A recent review discussed how noninvasive current stimulation over the visual cortex can affect visual perception (Bello et al., 2023). It suggested that anodal tDCS could marginally cause acute improvement in contrast perception, and offline current stimulation (including both tACS and anodal tDCS) could significantly enhance the ability of contrast detection. However, this study did not find any tES effect on response time, which contrasts our findings.
In a study that utilized the pedestal-delta-pedestal psychometric task to explore the contrast detection function of different visual pathways (Costa et al., 2015), a selective effect of tES was found. In the 4-alternate-forced-choice task, participants were asked to detect one of the four squares that increased (inferred parvocellular pathway) or decreased (inferred magnocellular pathway) its luminance in a different amount from other squares for a short period. Their results showed that during anodal tDCS, participants had a higher threshold for detecting the target of decreasing luminance than sham stimulation, indicating that anodal tDCS selectively affected the magnocellular pathway. The magnocellular pathway is associated with the processing of motion, depth, and high temporal and low spatial frequency information. In our study, the impact of polarity on brightness perception might also signify changes in the magnocellular pathway, as the perception of phosphenes inherently indicates the detection of temporal contrast of luminance.
Null result of intensity modulation on flash rate revealed independence of AM information coding in phosphene percept
Our flash rate scoring revealed a significant effect in the AM condition, consistent with our previous findings (Hsu et al., 2023). Participants perceived the flash rate as slower under AM stimulation than under sinusoidal stimulation. We observed a marginally significant interaction between intensity and polarity (p = 0.07), indicating that the flash rate was perceived faster under suprathreshold stimulation only in the case of anodal AM otDCS (p = 0.01). However, this effect was insignificant in other conditions, suggesting a limited overall impact of intensity on flash rate scoring. In essence, AM emerged as the primary determinant of the perceived flash rate, implying that the AM frequency may be encoded by the perceptual system and, at least partially, contribute to the perceived flash rate of phosphene.
The current results, therefore, suggest the constitution of the AM component in phosphene flash rate and provide strong evidence of detecting the temporal envelope despite the spatial envelope in visual perception.