Study participants

A total of 18 individuals, ages 17–48 years (15 males, 3 females), participated in the study. Participants were assigned to one of three groups (Table 1), each with six participants (5 males, 1 female): an LFA group, an HFA group, and a TD group. The diagnosis of autism was based on physician diagnoses, as confirmed by participants’ medical records and DSM-5 criteria [1], and scores of ≥10 on the Autism Diagnostic Observation Schedule Second Edition (ADOS-2) [8]. The mean age of participants with ASD was 29.8 years ±9.0; the mean age of participants with TD was 26.0 years ±4.8.

Table 1 Participant Demographics

Participants with ASD were classified as LFA or HFA based on the level of support required for daily living and on verbal communication abilities. Individuals classified as LFA required full support for daily living and were either nonverbal or minimally verbal; individuals with HFA required little or no support for daily living and demonstrated good verbal language abilities, including spontaneous speech production. To further assess language and cognitive function, all ASD participants completed the Peabody Picture Vocabulary Test and all but one ASD participant were also administered the Kaufman Brief Intelligence Test-2 (Table 1). The one exception was a participant with LFA (P1_LFA) who was unable to follow test directions to perform the Kaufman Brief Intelligence Test-2.

ASD participants were recruited through educational programs and a local institution that provides day services to individuals with autism and other developmental disorders. ASD and TD individuals were also recruited through newspaper advertisements and fliers posted at regional academic centers. Individuals were excluded from participation if they had a history of neurologic disease (e.g., epilepsy) or substance abuse, vision deficits, or demonstrated excessive movements or an inability to tolerate the electrode net used for the electrophysiology recordings. A total of nine individuals with LFA was originally recruited of whom three were excluded due to excessive movement and/or inability to wear the electrode net. All participants had normal or corrected-to-normal vision based on parental/caregiver reports (LFA) or self-reports (TD, HFA). Informed written consent was obtained from, or on behalf of, all participants: HFA and TD participants provided written informed consent; parents, guardians, or caregivers provided consent for LFA participants. All participants were paid $15 per hour for their participation. All procedures and consent processes were approved by the Institutional Review Board of the Johns Hopkins Medical Institutions.

Experimental procedures

Experimental set-up

The experimental set-up was the same for all participants. Each participant was tested individually in a quiet room where they were seated in front of a 34.5 x 27.5 cm LCD computer screen (1280 x 1024 pixel resolution, 60 Hz temporal resolution) with attached computer mouse. Viewing distance from the screen was approximately 50 cm; room lighting was dimmed to enhance visualization of the screen. Data collection was conducted by pairs of trained research assistants (HG, MM, LB) who were experienced in conducting electrophysiology studies with individuals with autism. A senior researcher (KS) with expertise in electrophysiology and signal processing was also present during the recording sessions to assist with the experimental set-up and technical issues and to monitor the recordings.

Based on prior observation that individuals with LFA often show increased agitation in unfamiliar situations, we implemented two familiarization strategies. First, all participants with LFA underwent one or two sessions of net tolerance training during the month prior to the study to familiarize them with wearing the electrode nets. Second, a family member or caretaker was present during testing, and at least one member of each researcher pair was introduced to the participant prior to the date of the study. The participant’s family member or caregiver also helped determine when session breaks were needed during testing.

Experimental stimuli

The main visual stimulus was a static black-and-white checkerboard pattern that appeared as a brief flash (100-ms duration) on the computer screen. The checkerboards were full-screen patterns comprised of 1.83 cm2 squares. A second visual stimulus was a pair of shapes, a white square and a white circle both 8 cm in size (length/diameter), presented side-by-side in the center of the screen. The location of each shape on the left or right side varied randomly across trials. The luminance of the white- and black-colored stimuli on the screen was 140 cd/m2 and 0.3 cd/m2, respectively.

Experimental task

Stimulus trials were presented consecutively with checkerboard trials interspersed randomly among shape-pair trials. The trial structure was the same for both stimuli. Each trial began with a blank (black) screen, lasting for 500 ms, followed by a white fixation cross in the middle of the screen for 1000 ms. This was followed by a second blank screen that lasted for 1200 ms, followed by presentation of either a checkerboard stimulus or a shape-pair stimulus. Each trial was followed by a 1000-ms blank screen. All stimuli were presented using E-prime software (version 2.0.8.90, Psychology Software Tools, Inc., Sharpsburg, Pennsylvania) running on a Windows XP PC (Dell OptiPlex 755, Dell Technologies, Round Rock, Texas).

The checkerboard trials were presented passively: no behavioral response was required and participants were not told about them beforehand. Passive presentation of the checkerboard trials controlled for individual differences in attentional states and behavioral performance across participants. The shape-pair trials served to encourage sustained visual attention to the computer screen by engaging participants in a behavioral task. Participants were instructed to respond on shape-pair trials by using the computer mouse to click on the circle shape. Participants with LFA were given verbal instructions, demonstrations, and practice trials before the session. Each pair of shapes remained on the screen until participants responded or until 10 s had elapsed. To ensure that participants moved the mouse to the target shape, the mouse pointer arrow returned to the bottom center of the screen before each trial. Participants’ behavioral responses were recorded online for analysis (accuracy, reaction time).

To minimize the time, the participants were required to sit still and to facilitate breaks during the session, and trials were grouped into five blocks of 108 trials each. Each block contained both visual stimuli in a 5:1 ratio of shape-pair trials to checkerboard trials. For two participants with LFA (P2 and P3), the number of stimulus blocks was increased to 10 (P2) and nine (P3) blocks of 72 trials. By both increasing the number of blocks and decreasing their length, participants could be given more frequent breaks while completing additional trials without extending the duration of the recording session.

Only checkerboard trials were included in the electrophysiology analysis. The total number of checkerboard trials presented to participants ranged from 108 to 140, depending on the number of stimulus blocks administered. For participants with LFA, the average number was 122 checkerboard trials. To verify that participants with LFA were attending to the computer screen when checkerboard stimuli were presented, two high-definition camcorders were used to video record each session: one recorded a frontal face view; the second recorded a rear view, including the computer screen (Sony HDR-CX360, Sony Corporation, NYC, New York).

Electrophysiology recording procedures

At the beginning of each recording session, a 256-channel HydroCel Geodesic Sensor electrode net that had been soaked in electrolyte solution (tap water with potassium chloride; 11,000mg/L) was fit onto each participant’s head. The electrode nets remained on for the entire session, including breaks.

Continuous EEG recordings were acquired at a sampling rate of 250 Hz, using a vertex electrode as the reference and an anti-aliasing filter cut-off frequency of 4 kHz. Electrode impedances were maintained below 50 kΩ. The recordings were acquired using a Geodesic EEG system, Net Amps 300 amplifier, and NetStation 4.3 software (Electrical Geodesics, Inc., Eugene, Oregon) on a Mac Pro computer (OS X version 10.6.8, Apple, Inc., Cupertino, California).

Each block of trials took approximately 9 min to complete. The session duration ranged from 45 min to 1 h, depending on the number of breaks required. For participants with LFA, several shorter sessions with longer breaks were scheduled on the same day to maintain adherence and increase the likelihood of acquiring recordings that were not contaminated by movement or other artifacts. The need for multiple shorter sessions was based on feedback from caregivers and informal evaluation of participants’ moods, behaviors, and willingness to follow directions. One participant with HFA returned for a second session because software difficulties resulted in early termination of the first session.

Data analysis

Electrophysiology recordings

Signal preprocessing

Signal preprocessing was performed using MATLAB (v. R2018b; Mathworks) and the EEGLAB toolbox [v. 14.1.0b; http://sccn.ucsd.edu/eeglab/;9]. The continuous EEG signals were high-pass filtered at 0.1 Hz and low-pass filtered at 38 Hz (Hamming windowed sinc FIR filter). Channels with excessive noise or artifact were identified for exclusion based on the voltage histogram [−500, 500 μV; bin size 10]. Channels with voltages that deviated from the mean-normalized voltage by ±30μV were excluded and interpolated with a spherical electrode configuration. The EEG signals were re-referenced to an average reference and segmented into epochs using a 2-s window that included a prestimulus period of 1000 ms. Epochs with ocular, cardiac, or other artifacts were identified for rejection based on independent component analysis (ICA) implemented in EEGLAB toolbox [10]. Because the data were relatively short in length, especially for participants with LFA, the principal component analysis was used to improve ICA outcomes by reducing the high-dimensional, 256-channel data to a smaller set of 64 principal components that accounted for ≥ 97% of the original data. The dimension-reduced data were then decomposed into independent components. The ICA components and their topographic distributions were examined to identify EEG artifact for rejection. Following ICA artifact rejection, only epochs (trials) associated with the checkerboard stimulus were selected for the visual evoked potential and power spectrum analysis. All checkerboard trials were reviewed to eliminate trials with excessive movement artifact, as determined visually (KS) and confirmed by review of the video recordings. Independent review of the video recordings was performed (HG, MM, LB) to eliminate any checkerboard trials that were presented when participants were not attending to the computer screen. The average trial rejection rates for each of the three participant groups were 21% for TD, 19% for HFA, and 30% for LFA participants.

Visual evoked potentials (VEP)

For each participant and electrode channel, VEP responses to the checkerboard stimuli were computed by trial averaging in the time domain. To focus on early cortical evoked potentials, the duration of the original 2-s epochs was trimmed to 700 ms [−200, 500 ms]. Each epoch (trial) was then normalized to the prestimulus voltage by subtracting the mean prestimulus voltage [−200, 0 ms] from the poststimulus signal. VEPs were measured at the occipital midline electrode (Oz), where the largest VEPs and alpha oscillations were observed across all participants.

For each participant, peak response latency and amplitude measurements (base-to-peak) were derived for the P1-N1-P2 components in the 0–300 ms period following stimulus presentation. The P1 was identified as the largest positive waveform deflection between 45 and 70 ms after stimulus presentation; the N1 was the largest negative waveform following the P1, occurring between 70 and 130 ms; and the P2 was the next largest positive waveform occurring between 120 and 250 ms. The total number of checkerboard trials averaged for group VEPs was 494 (LFA), 407 (HFA), and 397 (TD). The average number of checkerboard trials for each group was 82.3±18.3 (LFA), 67.8±10.9 (HFA), and 66.2±8.2 (TD). Group-level VEPs were computed by averaging across participants within a group, and peak waveform measurements were derived using the same procedures implemented for individual waveform measurements.

Spectral power analysis

Time-frequency analysis was used to measure the overall power of cortical EEG oscillations and to identify event-related changes in spectral power (ERSP) with visual stimulation. Time-frequency analysis of single trials was performed using wavelet analysis (newtimef function, EEGLAB). To ensure the response time window was sufficiently long to capture the slower cortical EEG oscillations (<10 Hz), the original, 2-s epoched data was used. Time-frequency analyses were performed at both the individual and group levels.

To compute the overall time-frequency power spectrum, the power of 27 linearly spaced EEG frequencies from 4 to 30Hz was calculated using a wavelet time window of 1 s. This was done for all 256 channels to generate cortical topographical maps of the power distributions. The time window of the resulting power spectrum was then trimmed to −500 to 500 ms. The overall time-frequency power spectrum was plotted along with the log-transformed power spectrum density profile for the prestimulus and poststimulus periods. To further examine the cortical distribution of spectral power, topographical maps of the pre- and poststimulus spectral power were plotted across all electrode channels with interpolation (topoplot function, EEGLAB). ERSP was calculated for channel Oz by subtracting the mean prestimulus power of each frequency from the corresponding frequency occurring 0–500-ms poststimulus. We used non-parametric, Monte Carlo-based permutation testing methods in EEGLAB to assess the statistical significance of poststimulus changes in spectral power relative to baseline (prestimulus) at an alpha level of 0.05. To correct for multiple comparisons that can inflate type I error rates, we used the false discovery rate [11]. We also performed the power spectrum analysis at the individual level to compute bootstrapped confidence intervals for alpha and theta powers of prestimulus and poststimulus time windows.

Behavioral analysis

Although shape-pair trials were not included in the electrophysiology analysis, we analyzed participants’ behavioral response accuracy and latency to confirm comprehension of the task, adherence, and visual attention to the computer screen.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Disclaimer:

This article is autogenerated using RSS feeds and has not been created or edited by OA JF.

Click here for Source link (https://www.biomedcentral.com/)