Event Related Potentials (ERP) Facility
When brain cells fire, they release tiny amounts of electricity that travel through the brain and scalp to escape into the outside world. This bid for freedom is very convenient because it allows us to measure a person's brain responses from outside their head. We do this using small sensors that are sewn into a soft cap.
At the ERP Faciltiy, we measure two types of brain responses: event-related potentials (ERPs) and fixation-related potentials (FRPs). An ERP is the average brain response to a particular stimulus, such as a high-pitched musical note or the photo of a face. An FRP is the average brain response to one part of a whole visual stimulus, such as the eyes in a face, or a word written in a sentence.
People and Projects
I am investigating how we can improve the methods of functional Transcranial Doppler Ultrasound (fTCD) experiments looking at the lateralisation of language. In the current word generation paradigm, participants are instructed to 'relax' following a trial. A problem with this approach is that we don't really know what a person is thinking of when they are told to 'relax'. In my study, we examined the time taken for blood flow velocity of the left and right middle cerebral arteries to normalise in the 'relax' condition, compared to an alternative 'counting' condition. By instructing a person to count, we hope that normalisation will occur sooner. In principle, shorter trials could make it easier to use fTCD on difficult to test populations, such as young children. In addition, this approach might improve the reliability of fTCD for word generation tasks.
In collaboration with Nic Badcock, Betty Mousikou tested the validity of Emotiv EPOC® ERPs against research-grade Neuroscan ERPs. Recent advances in brain-computer input devices have seen a number of companies develop EEG systems for gamers. These systems make actions on a computer based on patterns of EEG activity. What's more, they tend to be wireless, easy to setup, and much cheaper than research EEG systems. The outcomes of the validation experiment indicated that the Emotiv EPOC and Neuroscan systems produce similar reliable auditory ERPs.
Dr Florian Hutzler and Stefan Hawelka
In July 2008, Florian and Stefan travelled from the University of Salzburg (Austria) to MACCS (Australia) to integrate an eye-tracking system with our existing ERP lab so that we can measure fixation-related potentials (FRPs). FRPs allow us to measure ERPs to stimuli presented in real-world contexts. There are only 5 active FRP labs in the world. This Austrian-Australia axis are collaborating on 3 experiments: one comparing FRPs to words in Austrian and Australian readers, and two comparing FRPs to eyes and mouths in faces.
Professor Genevieve McArthur
We tested 6- to 12-year-old children with SLI (N = 19), children with SRD (N = 55), and controls (N = 36) for their passive auditory ERPs to tones, rapid tones, vowels, and consonant-vowels. Thirty-eight percent of children with SLI or SRD had atypical passive auditory ERPs in the N1-P2 window across all the sounds. These children had atypically ERPs in the N1-P2 region, as well as poor nonword reading and poor nonword repetition. These findings support the idea that impaired auditory processing is a cause of SLI and SRD.
Associate Professor Greg Savage
Verbal memory difficulties due to left hemisphere lesions are reliably measured using word-based tests; nonverbal memory problems due to right hemisphere lesions are measured with pictures, but less reliably. We tested hemispheric biases using ERPs to (1) written nonwords, spoken nonwords, and the two paired; and (2) dot patterns, novel melodies, and the two paired. Temporal analyses revealed a larger N1 to written nonwords on the left and dot patterns on the right. Spectral analyses will be used to measure later neural processing.
Together with Nic Badcock, Hannah our study explored the relationship between reading ability and language lateralisation while performing easy and hard language tasks. Language lateralisation was estimated using functional Transcranial Doppler Ultrasound. The study found that when the task was easy, there was no significant relationship between reading ability and language lateralisation. However, when the task demands increased, poor reading was associated with increased right lateralisation and good reading was associated with increased left lateralisation. Insight into the brain regions implicated in poor reading may have practical implications for the development of reading interventions that attempt to directly alter brain function.
After a miraculous recovery from a Death Eater blast, Professor Hedwig has joined the ERP-FRP lab to help explain to children how ERPs work so that they will not be frightened of the sensors that we put on their heads. Professor Hedwig has helped us test 100's of children, making her the most experienced ERP researcher in the lab. She has also been photographed with more fans (in their ERP caps) than Paris Hilton. She is currently pursuing a collaboration with Kermit-the-Frog in the Language Acquisition Lab.
Word Generation is the gold standard paradigm used to determine which brain hemisphere dominates while performing a language task. A visual preparatory cue 'clear mind' is used within this paradigm to focus attention on preparation for silent word generation. While an auditory cueing tone has been experimentally compared to no cue, the influence of 'clear mind' had not been examined and became the focus of my study. It was found that the cue 'clear mind' was more internally reliable for assessing hemisphere dominance for language than a cue containing '####' symbols.
I compared the performance of monolinguals and bilinguals on the Attentional Network Task. I found that that bilinguals were better at resolving response conflict than monolinguals, and exhibited a smaller P300 in the left hemisphere at P3 and CP3. In contrast, I found that bilinguals and monolinguals showed comparable spatial attention as indexed by N1 and P1, and comparable temporal attention as indexed by a sustained negative deflection (CNV) over frontal sites.
This study examined the process of switching associations between facial identity (e.g., person 1 and person 2) and different emotional expressions (i.e., happy and angry). ERPs and behavioural measures were measured while subjects monitored facial expressions of identity pairs. Behavioural responses were more impaired when they had to switch their behavioural responses to angry than to happy expressions. Switching associations between facial identity and emotional expression also modulated the N2, P3a and P3b ERP components.
Mridula Sharma and Krystal Lo
Behavioural studies have shown that adding a visual cue to an auditory cue makes the latter appear louder. However, behavioural data are affected by task-related factors such as attention. ERPs can be measured without task-related factors. We aim is to use ERPs to examine the auditory-visual advantage in 12 adults while they perceive visual and auditory interaction of congruent stimuli (the subject sees a person voicing /ba/ and hears /ba/) and incongruent stimuli (the subject hears /ba/ but the person voices /ga/).
I am interested in how our brains help us coordinate with other people during social interactions to share a common focus of attention. This cognitive ability is called joint attention, and is believed to be pivotal for language development and understanding the mental states of others. My ERP study investigates the time course of brain activity associated with evaluating the social relevance of gaze, which is an essential component of joint attention. I have developed a virtual reality paradigm, where this process can be investigated in a truly interactive context; enhancing ecological validity whilst maintaining experimental control.
Together with Betty Mousikou and a team of researchers, Nic Badcock has tested the validity of a gaming EEG system (Emotiv EPOC®) for research quality ERP measurement. The team adjusted the Emotiv EPOC® system so it could be used to measure auditory ERPs, and then compared these ERPs to auditory ERPs measured with a research Neuroscan system. There was a good match between reliable auditory ERPs in the two systems in adults. His next step is to validate Emotiv EPOC® ERPs in children. And, pending funding, seals.
Peter De Lissa
My ERP study tested facial expression processing in adults and children. As part of a broader study conducted by Romina Palermo at the MACCS ERP lab, I used ERPs to measure cortical activity in two different age groups, using participant sex as a variable of interest. No sex-effect was found to be significant in either age group. However, different emotional expressions elicited quantitatively and qualitatively different cortical activations in the temporo-parietal regions (generally regarded as reflecting activation of the fusiform face area) as well as in more frontal areas at a later latency.
Dr Romina Palermo
A complex network of brain regions are involved in perceiving and recognising facial expressions. Some brain pathways and structures appear to process faces relative automatically while others seem to be involved in conscious recognition of expressions. It is likely that the brain regions involved in automatic face processing mature earlier than those involved in conscious processes. This research aims to chart the development of automatic and conscious processing of facial expressions to see if these process develop differently with age.
I measured ERPs to happy faces (judged to be attractive), sad faces (judged to be unattractive), and neutral faces. Unattractive sad faces and neutral faces triggered larger P1 and P300 ERPs than attractive happy faces, indicating a processing bias towards unattractive faces. Happy attractive and sad unattractive faces triggered a larger N170 ERP than neutral faces. This suggested that the processing of attractiveness was enhanced by happy and sad expressions.
Together with Nic Badcock I am conducting a functional Transcranial Doppler ultrasonography (fTCD) project to investigate lateralisation of word production. fTCD is a quick and non-invasive method that can be used to compare blood flow velocity in the left and right cerebral arteries to investigate cerebral lateralisation of a cognitive process, such as language production.
The word generation paradigm (e.g., Knecht et al., 1996, Neuroreport) has been widely used to investigate lateralisation of language production. During this task a participant is asked to generate words starting with a letter presented on a screen. A large number of studies using this paradigm have reproduced a left hemisphere lateralisation in most individuals. However, it is unclear how subtle behavioural differences such as the number of words produced, influences the reliability of the methodology. In the current experiment we are testing whether the number of words produced matters.
Prosody in speech and music is cued mainly by changes in duration. The study investigated the relationship between neural processing of sound duration in speech and music. Mismatch negativity (MMN) was recorded for duration changes in tones and abstract duration changes in speech and music. The results indicated a relationship between abstract MMN for music and MMN for duration deviance in tones. However, there was no relationship between abstract MMN for speech and abstract MMN for music.
There is considerable controversy about how sound and visual processing matures in the brain. Research has revealed that adolescence is a period of great neural change. The current research is aimed towards studying the neurophysiological changes associated with auditory and visual processing during adolescence. Using ERPs development changes to speech and non speech sounds and visual images will be tracked from 10-18 years.
Developing objective measures of listening effort during speech processing
A group of researchers from the Department of Linguistics (Kelly Miles, Chi Lo, Catherine McMahon, Isabelle Boisvert and Ronny Ibrahim) and the ARC Centre of Excellence in Cognition and its Disorders (Peter de Lissa) are using event-related potentials to develop objective measures of listening effort during speech recognition tasks in adults with cochlear implants and adults with normal hearing. They are also examining the relationship between slow frequency oscillations, pupil size, and speech recognition performance in adults with normal hearing.
Emotiv EPOC validation
In 2010, Gen McArthur had the idea that perhaps a newly developed, commercially available, gaming EEG system might be useful for research. It's quick to set up, portable, the data saves wirelessly, and it's significantly cheaper than research-grade EEG systems. So, we've put some effort into figuring out how Emotiv EPOC® compares to the research-grade Neuroscan EEG system.
Our strategy was to simultaneously record EEG using the two systems. We put a few extra holes in our EasyCaps, set people up with Neuroscan, and then put Emotiv over the top. We've finished three experiments now, auditory odd-ball paradigms with adults and children, and a visual n170 face paradigm with adults. Long story short, the two systems record similar data.
Investigating the Effects of tDCS on Reading Performance
Transcranial direct current stimulation (tDCS) is a cutting edge non-invasive brain stimulation technique used in cognitive neuroscience. tDCS generates an electromagnetic field in the brain that either enhances (anodal) or inhibits (cathodal) neuronal activity. tDCS allows us to study whether stimulating the brain enhance mental processes such as learning, and whether this is influenced by the intensity, duration, location and polarity (i.e. anodal or cathodal) of the stimulation. During tDCS, participants wear a cap or headset fitted with electrodes that deliver current to the scalp. The stimulus is comparable to electrical currents that are naturally produced by the brain – it is completely silent, safe and portable.
We are currently investigating whether tDCS can enhance reading performance in adults with typical reading ability. We also plan to apply this technique in a full scale clinical trial of reading training involving adults with dyslexia, to determine whether tDCS can facilitate the effects of the reading training intervention.
Exploring the role of meditation in language remediation: A behavioural and neurophysiological study
A cognitive mechanism fundamental to early spoken language processing - auditory attention - has been found to be impaired in people with aphasia, dyslexia and ADHD. Previous research has shown that meditation can improve behavioural and neurophysiological indices of auditory attention in typical adults.
From 2011 to 2014, our team, incorporating researchers from the Department of Cognitive Science and collaborators from various Research Programs within the ARC Centre of Excellence in Cognition and its Disorders (CCD), have assessed changes in auditory attention with ERPs and behavioural measures in a meditative and non-meditative condition in 16 long-term meditators (including Buddhist nuns and monks) and 16 non-meditators in an unfunded pilot project.
In 2014, we were awarded funding under the CCD's Cross Program Support Schemewhich will be used to implement the next phase of our program. This will use a randomised controlled trial to compare the effects of 8-weeks of meditation training (experimental group) to a control group on ERP and behavioural measures of auditory attention, memory, and spoken language.
Information for New Users
ERP and FRP experiments should not be take on lightly. They take longer to set up, measure, analyse, and interpret than behavioural responses. if you are thinking of doing an ERP or FRP experiment, the first thing you need to do is ask yourself: "can I answer my question using behavioural responses". If the answer to this is yes, then do not do an ERP or FRP study. If the answer is no, then read on.
If you have never done an ERP or FRP experiment before, you will need both theoretical (i.e., how to set up your experiment and how to interpret the data) and practical support (how to measure ERPs and FRPs). The ERP Lab can provide practical support but not necessarily theoretical support. Thus, anyone who wants to do an ERP or FRP experiment must have a researcher on the project who is an ERP or FRP expert in the relevant field of research. If you are a PhD student, this expert must be on your supervisory panel.
Once you have designed your experiment with the expert on your research team, you need to arrange a meeting with the executive committee to discuss (1) if the lab has the equipment to support your experiment, (2) when the lab is free for you to run the experiment, (3) when you should do the training course. You will also need to discuss funding. Researchers with external funding, and PhD students with Cognitive Science funding, are asked to pay $20 per subject for consumable and wear-and-tear expenses. Alternative arrangements will be made with researchers and students without funding.
New researchers to the ERP facility - both experienced and inexperienced - need to do a training course before they can start testing in the lab. The first step is to have an introductory session a member of the ERP Facility Committee. The second step is to run your entire experiment on yourself. The third step is to run your experiment on a number of (forgiving) friends. The fourth step is to analyse their data and see if it makes sense to proceed. If so, you need to arrange two testing sessions with two members of the ERP Facility Committee. If you pass both tests sessions, then you will be "signed off", and you will be free to run your experiment unsupervised. All these steps are outlined in this Training Certificate.
We understand that this process appears very dictatorial. However, the top priority in the ERP facility is the subject - it is not the experiment or the experimenter. Thus, we need to make sure that every researcher in the ERP facility follows our subject-friendly procedures.
Neuroscan ERP System
We use a 32-channel SYNAMPSII amplifier from Neuroscan, which measures ERPs at all frequencies, including brainstem responses. This system is suitable for high-density, low noise recordings. Our headbox supports 70 channels: 64 monopolar, 4 bipolar, and 2 high-level channels. Participants' electroencephalograms (EEGs) are detected at the scalp using Ag/AgCl sintered electrodes that are held in place by Quik-Caps (small, medium, and large). The EEGs are recorded using SCAN Acquire software and processed to produce ERPs using the SCAN Analysis software.
The stimuli are presented using Presentation® software that drives a SoundBlaster Audigy2 ZS soundcard and NVIDIA GEForce FX 5200 videocard. Sounds are presented through speakers and Sennheiser headphones. Images are presented on a 19-inch CRT Monitor. The resolution of the system (i.e. the time delay between when a stimulus code is recorded on the EEG and when the stimulus is actually presented) is less than 1 ms.
An FRP is the average brain response elicited when natural eye-fixations are made to visual stimuli, such as the mouth or eyes in a face, or a word written in a sentence. We measure FRPs using a method developed by our collaborators in Salzburg. In this system an eye-tracker monitors and records eye-movements, and sends codes to the ERP system at the beginning of a stimulus presentation, or when the eyes fixate on a pre-defined area of a visual scene for more than 35 ms. We use these codes and the record of eye-movements to calculate the average brain activity in response to the presentation of a visual stimulus, as well as subsequent fixations made to one or more areas of the stimulus. For example, we can calculate the average brain response to the eyes in a face and the average brain response to the mouth in the same face, by directing initial fixations to be over specific regions of a stimulus when presented. The full FRP paradigm then allows us to record the brain activity that occurs when participants' eyes then fixate on various parts of the face during natural scanning. This allows us to observe brain activity during dynamic visual processing, and is a very useful technique if you want to measure brain responses to stimuli presented in real-world situations such as words within sentences, eyes in faces, or (perhaps) musical notes in scores.
Emotiv EPOC EEG/ERP System
We use an Emotiv EPOC® EEG system that was designed as a brain-computer-interface device. We have validated this system against Neuroscan for measuring event related potentials to auditory and visual stimuli and it produces equivalent results. The system is limited with respect to the electrode layout and sampling rate but has the benefit of being light-weight, easy to setup, and portable.
The EEG electrodes are housed within a one-sized fits all (we haven’t tested below age 6) headset. The EEG sensors are gold-plated and contact is made using saline soaked cotton rolls (as opposed to the gel we use for Neuroscan). The headset had 16 sites, aligned with the 10-20 system: AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8, FC4, M1/P3 and M2/P4. We use the mastoids (M1 and M2) as reference sites (thought P3 and P4 can also be used). M1 acts as a ground reference point for measuring the voltage of the other sensors, while M2 acts as a feed-forward reference point for reducing electrical interference from external sources. Only signals from the other 14 non-reference channels are saved at a sampling rate of 128Hz.
EyeLink 1000 tower-mounted system
The tower-mounted infar-red EyeLink 1000 eye-tracking system by SR Research provides the greatest spatial and temporal resolution. It operates at 2000Hz (monocular), and is capable of sampling eye-position from corneal reflection, as well as pupillometry data. The participant's head is stabilized on a chin-rest in a tower. The camera is positioned above the chinrest and monitors their eye-movements via a mirror placed between the participant and the computer screen. The mirror reflects infra-red light but visible light passes through, so the participant’s view of the computer screen is not affected. Our tower-mounted eye-tracker is integrated with a Neuroscan EEG system, allowing brain responses to be time-locked to fixation on particular areas of interest on the screen.
Functional Transcranial Doppler Ultrasound System
The lab hosts a DWL Doppler Box system with bilateral monitoring 2MHz ultrasounds probes, which can be attached to either the DiaMon DWL headset or Spencer Technology’s Marc600 headframe. DWL’s QL software is used for data recording and we use the ‘dopOSCCI’ MATLAB toolbox for data processing (written by one of our lab managers).
Functional Transcranial Doppler Ultrasound (fTCD) is a non-invasive imaging technique that takes advantage of Doppler physics to determine the rate of blood flow to the brain, often within the middle cerebral arteries. By simultaneously examining the rate of blood flow in response to some cognitive task (e.g., language or spatial processing), fTCD can be used to determine whether the neural substrates for the cognitive process are predominantly located in the left or right hemispheres of the brain; i.e., the lateralization of cognitive processes.
Although the spatial resolution of fTCD is very limited compared with magnetic resonance imaging (MRI), simply reflecting the neuro-metabolic coupling to very general regions of the cortex (i.e., the middle cerebral arteries supply blood to approximately 70% of the brain), the temporal resolution is high, sampling the rate of blood for 100 times per second (100 Hz). Additional advantages of fTCD are that it is inexpensive to use, portable, and it is robust to movements which would spoil MRI recordings. Therefore different experimental paradigms are feasible with fTCD and it is suitable for use with very young children.