[Eeglablist] Single trial analysis

Tarik S Bel-Bahar tarikbelbahar at gmail.com
Wed Oct 11 12:45:04 PDT 2017

Hello Gian, some quick notes below, best wishes!


There are several papers out there on single-trial metrics, both in
mainstream EEG journals, imaging journals, methods journals, as well as
IEEE/BCI/neurorehab/neuro-engineering journals. This is an important topic
as the field ihas been moving away (more quickly in recent years) from
"averaged" responses to understanding variability across trials, the
validity of using single trials, etc.. Take a good look through those
publications for good ideas. If you have trouble finding those, just look
for any articles in last 10 years that mention single-trial and EEG in
their title or abstract, and review them all. Then you can become an expert
on what's been done, what's not been done, and contribute to this area of
EEG methods/metrics. Note that the BCI field has also dealt extensively
with working with single-trials and there are great reviews out there
summarizing the efficacy of those methods, and the accuracy of various
machine/classifier approaches to working with BCI-related protocols.

I've include a range of mostly recent papers (including several reviews
focused on the single-trial topic) below. You can find them all easily on
Google Scholar. Note there are many other papers that I am not listing
here. When you have developed a good list of ideas/methods/etc... please
consider sharing them with the list as it will also help other users make
benefit of the single-trial brain dynamics, and make informed decisions.

Within eeglab, you can certainly pull out single trial metrics, and you
already have important issues/caveats in mind. As you know eeglab
emphasizes the power and usefulness of ICA, so one thing you can do is stay
in ICA space, use good neural IC components, and focus on IC-resolved
single-trial metrics. In fact the developers of eeglab had a strong focus
on making it easier to visualize and work with single-trial brain dynamics,
as there's where some of the most interesting "action" happens. In other
words, humans naturally function in terms of single-trials, it's only
science that's needed to start with averaged responses (though, again, this
has been changing for at least the last two decades). When you start your
trials of single-trial analyses with eeglab, you can send more focused and
specific questions to the list, especially if you bump into problems of
implementation in terms of the eeglab gui and scripting your methods.

I've also seen some paper that essentially drop all ICs except the one of
interest, and then looks at the rebuilt channel-level data based on that.
This is one way of many to "single-out" the activity of interest at each
trial. Overall, you would want to pull/use single-trials that really match
some prototype ERP, but you would have to deal with the fact that ERPs are
usually averaged across trials. Explore the ERPimage technique in eeglab at
least to get a sense of how you can visualize data across all your trials,
including sorting them in various ways. Another set of issues one would
have to deal with is cleanliness of the data, method of preprocessing,
individual differences in signal quality, variation in ERPs across single
trials within each participant, and the validity of the protocol you have
used to generate the ERPs in question, including the quality of your
specific implementation of that protocol.

Best practice for single-trial detection of event-related potentials:
Application to brain-computer interfaces

Simultaneous recording of MEG, EEG and intracerebral EEG during visual
stimulation: from feasibility to single-trial analysis

The effects of reward magnitude on reward processing: An averaged and
single trial event-related potential study

Exploiting the intra-subject latency variability from single-trial
event-related potentials in the P3 time range: A review and comparative
evaluation of methods

Similar sound intensity dependence of the N1 and P2 components of the
auditory ERP: averaged and single trial evidence

Single-trial analysis and classification of ERP components—a tutorial

Single-trial normalization for event-related spectral decomposition reduces
sensitivity to noisy trials

A Bayesian method to estimate single-trial event-related potentials with
application to the study of the P300 variability

Functional source separation improves the quality of single trial visual
evoked potentials recorded during concurrent EEG-fMRI

The superior temporal sulcus and the N170 during face processing: single
trialanalysis of concurrent EEG–fMRI

Attentional selection in a cocktail party environment can be decoded from
single-trial EEG

Space-by-time decomposition for single-trial decoding of M/EEG activity

Neural Activity Elicited by a Cognitive Task can be Detected in
Single-Trials with Simultaneous Intracerebral EEG-fMRI Recordings

A Change-point Analysis Method for Single-trial Study of Simultaneous
EEG-fMRI of Auditory/Visual Oddball Task

Single-trial analysis of readiness potentials for lower limb exoskeleton

How attention influences perceptual decision making: Single-trial
EEGcorrelates of drift-diffusion model parameters

Predicting EEG single trial responses with simultaneous fMRI and relevance
vector machine regression

Individual differences in human auditory processing: Insights from
single-trialauditory midbrain activity in an animal model

A pipeline of spatio-temporal filtering for predicting the laterality of
self-initiated fine movements from single trial readiness potentials

Single-trial detection of somatosensory evoked potentials by probabilistic
independent component analysis and wavelet filtering

Selectivity of N170 for visual words in the right hemisphere: Evidence
fromsingle‐trial analysis

Single-trial event-related potential extraction through one-unit

A single trial analysis of EEG in recognition memory: Tracking the neural
correlates of memory strength

Interpretable deep neural networks for single-trial EEG classification

Machine learning for real-time single-trial EEG-analysis: from
brain–computer interfacing to mental state monitoring

Exploring time-and frequency-dependent functional connectivity and brain
networks during deception with single-trial event-related potentials

Trial‐by‐trial co‐variation of pre‐stimulus EEG alpha power and
visuospatial bias reflects a mixture of stochastic and deterministic effects

The Pavlovian craver: Neural and experiential correlates of single
trialnaturalistic food conditioning in humans

Single-trial EEG classification of motor imagery using deep convolutional
neural networks

Open Access Dataset for EEG+ NIRS Single-Trial Classification

Pre-trial EEG-based single-trial motor performance prediction to enhance
neuroergonomics for a hand force task

Optimal spatial filtering of single trial EEG during imagined hand movement

Using single-trial EEG to predict and analyze subsequent memory

Classifying single trial EEG: Towards brain computer interfacing

A comparison of single-trial EEG classification and EEG-informed fMRI
across three MR compatible EEG recording systems

Toward FRP-Based Brain-Machine Interfaces—Single-Trial Classification of
Fixation-Related Potentials

Spatio-spectral filters for improving the classification of single trial EEG

Sliding HDCA: single-trial EEG classification to overcome and quantify
temporal variability

Spatiotemporal representations of rapid visual target detection: a
single-trial eegclassification algorithm

The P600-as-P3 hypothesis revisited: Single-trial analyses reveal that the
late EEG positivity following linguistically deviant material is reaction
time aligned

Single Trial EEG Patterns for the Prediction of Individual Differences in
Fluid Intelligence

Single-trial EEG classification using logistic regression based on ensemble

RSTFC: A novel algorithm for spatio-temporal filtering and classification
ofsingle-trial EEG

Neurological classifier committee based on artificial neural networks and
support vector machine for single-trial EEG signal decoding
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20171011/880b396f/attachment.html>

More information about the eeglablist mailing list