Peak Selection and Latency Jitter Correction in Developmental ERP

• In Event: Poster Session 07
  In Poster Session: PS 07 Topic - Biological Processes: Neuroscience and Genetics

Abstract

Event-related potentials (ERPs) provide great insight into the timing of neural responses to discreet stimuli. However, developmental ERP work is plagued with inconsistent approaches to identifying and quantifying ERP latency, which lead to unreliable results. There is a popular practice in the field to select ERP component time windows a priori based on previous research. However, latencies of ERP components have been shown to change as the brain matures, especially in infants and young children, thus a priori time window selection may lead to undesirable variability in results across studies. Additionally, individual differences in the timing of neural responses can increase variability in results observed within an experimental group. Data-driven analysis strategies may provide a more accurate picture of one’s specific dataset. In the current study, the effects of method of peak selection on infant ERP visualization and analysis were examined in data previously collected from 12-month-old infants. Participants completed an ERP study including visual presentations of intact and scrambled face and toy stimuli. Data was analyzed using three approaches for component time window selection including 1) the a priori method for peak selection, 2) a method developed to define component latency in individual participant’s ERP averages, and 3) a method aimed at identifying ERP component latencies in individual trials. For the second approach, a semi-automatized procedure was used to define component latencies in individual ERP averages; this method has been used in our previous work with infants (Conte et al., 2020; Guy et al., 2016, 2018). A custom MATLAB script was used to identify the point of greatest amplitude for each stimulus condition at specific electrodes within a predetermined time window. Each peak was inspected after detection and misidentified peaks were adjusted. This has helped to overcome weaknesses of a “one size fits all” approach to peak selection, however it does not account for trial-by-trial variability within a participant. This issue, known as ERP latency jitter, may blur the average ERP, misleading the interpretation of neural mechanisms. In our third approach, a further step in the data pre-processing pipeline was taken to estimate and correct latency jitter. The freely available ReSynch MATLAB toolbox (Ouyang, 2020) was used to determine whether component latency jitter was present in individual trials at electrodes of interest, correct for jitter, and then average jitter-corrected trials. The preliminary results of our study indicate that the use of individual peak selection and the ReSync toolbox provide better characterization of individual participant (Figure 1a) and group data (Figure 1b). Thus, the implementation of these pre-processing procedures is expected to lead to more precise peak selection in components characterized by distinct peaks (e.g, the P1, N290), but may be unsuitable for examination of ERP components characterized by broader, less precise peaks (e.g., the Nc).

Authors

• Maggie W Guy, Loyola University Chicago
  Presenting Author

• Aslı Bursalıoğlu, Loyola University Chicago
  Non-Presenting Author

• Stefania Conte, University of South Carolina
  Non-Presenting Author