Effects of spectral smearing of stimuli on the performance of auditory steady-state response-based brain–computer interface

Abstract

There have been few reports that investigated the effects of the degree and pattern of a spectral smearing of stimuli due to deteriorated hearing ability on the performance of auditory brain-computer interface (BCI) systems. In this study, we assumed that such spectral smearing of stimuli may affect the performance of an auditory steady-state response (ASSR)-based BCI system and performed subjective experiments using 10 normal-hearing subjects to verify this assumption. We constructed smearing-reflected stimuli using an 8-channel vocoder with moderate and severe hearing loss setups and, using these stimuli, performed subjective concentration tests with three symmetric and six asymmetric smearing patterns while recording electroencephalogram signals. Then, 56 ratio features were calculated from the recorded signals, and the accuracies of the BCI selections were calculated and compared. Experimental results demonstrated that (1) applying smearing-reflected stimuli decreases the performance of an ASSR-based auditory BCI system, and (2) such negative effects can be reduced by adjusting the feature settings of the BCI algorithm on the basis of results acquired a posteriori. These results imply that by fine-tuning the feature settings of the BCI algorithm according to the degree and pattern of hearing ability deterioration of the recipient, the clinical benefits of a BCI system can be improved.

Effects of degree and symmetricity of bilateral spectral smearing, carrier frequency, and subject sex on amplitude of evoked auditory steady-state response signal

“…Second, we already verified the effect of variations in the smearing parameter of the simulator in a previous report, which showed that the auditory filter bandwidth of the simulator output widens as the value of the smearing parameter increases (Hwang et al 2015). To find the relevant values of the smearing parameter for each HI setup, we referred to a report of Glasberg and Moore (1986) that demonstrated that the equivalent rectangular bandwidth increased monotonically as the absolute threshold increased, as in our recent report (Hwang et al 2017). In the Glasberg study, the ratios of the equivalent rectangular bandwidth (ERBR) between HI and NH for a 1000 Hz pure-tone input signal were approximately 1.43, 2.31, and 3.18 for ML, MSL, and SL individuals, respectively.…”

“…As we mentioned in the Introduction, the effect of the former is not serious in BCI applications because it can be avoided by selecting proper frequencies of auditory stimuli and adjusting the stimuli volume to a comfortable level for the individual; however, avoiding the effect of the latter is not a simple problem. As we demonstrated in our previous report, the degree and pattern of spectral smearing can affect the performance of an ASSR-based BCI system because they can affect the perception of the stimulus, the degree of concentration on the stimulus, and comfort in concentration (Hwang et al 2017). For this reason, to improve the clinical usability of the ASSR-based BCI system for persons with sensorineural HI, it is important to reflect the degree and pattern of spectral smearing of the individual in the parameter fine-tuning.…”

“…Based on those measurements, we set the values of the smearing parameter as 0.5 (ERBR = 1.43), 1.0 (ERBR = 2.15), and 1.5 (ERBR = 3.28) for the ML, MSL, and SL setups, respectively. Then, after verifying the relevance of the utilized simulator, first, two sinusoidal amplitude-modulated stimuli-S 500 {37} and S 2000 {43} (also denoted by S 500 {37} NH and S 2000 {43} NH )-were synthesized with 100% modulation depth, where 500/2000 represented the carrier frequency in Hz and 37/43 represented the modulation frequency in Hz, respectively (Hwang et al 2017;Kim et al 2011). Next, to synthesize the spectral smearingreflected stimuli, S 500 {37} NH and S 2000 {43} NH were individually applied to the simulator with ML, MSL and SL setups.…”

“…In this study, we assumed that if the ratio was C 2, the ASSR was generated, and if the ratio was \ 2, the ASSR was not generated. Based on these criteria, only the trials that contained the proper ASSR signal were utilized for postexperimental statistical investigation (Hwang et al 2017). In addition, the ASSR phase was neglected during the investigation based on a report of Gander et al, which demonstrated that attention had no detectable effect on the ASSR phase (Gander et al 2010).…”

“…In our previous study, we demonstrated that the application of spectral smearing to auditory stimuli can decrease the performance of ASSRbased BCI algorithms. To do this, we used an eight-channel vocoder and 10 individuals with NH (Hwang et al 2017). However, in that study, we divided the severity of spectral smearing into only three classes (normal, moderate, and severe) and did not investigate the effect of sex, because we recruited only three female participants.…”