Estimates one deterministic artifact subspace across trials and reconstructs the retained activity through a channel-matched lead field.
Usage
tmsSSPSIR(
x,
forward_model,
artifact_window_ms = c(5, 50),
n_artifact = NULL,
variance_fraction = 0.9,
lambda = 0.05,
reference = "average",
epoch_start_ms = 0,
assay_name = NULL,
output_assay = "tms_sspsir",
overwrite = FALSE
)Arguments
- x
A pulse-cleaned
PhysioExperiment.- forward_model
A list containing a finite channel-matched
leadfield.- artifact_window_ms
Closed artifact-training interval.
- n_artifact
Optional exact number of artifact components.
- variance_fraction
Cumulative artifact variance target when
n_artifactisNULL.- lambda
Positive source-reconstruction regularization.
- reference
Currently exactly
"average".- epoch_start_ms
Time of the first epoch sample relative to the pulse.
- assay_name
Optional pulse-cleaned input assay.
- output_assay
Output assay name.
- overwrite
Whether to overwrite the selected input assay.
Details
The pulse must already be removed or hardware absence must be explicitly
declared in metadata(x)$tms_eeg$pulse_absent. Both data and lead field are
average referenced with the same matrix. The returned data use the fixed
source-informed reconstruction Lr G P, not the plain SSP projection P.
SSP-SIR can attenuate neural signal when the artifact and neural subspaces
overlap; a lead field is an approximate physical model, not proof that a
retained deflection is cortical.
References
Mutanen TP, Kukkonen M, Nieminen JO, et al. (2016). Recovering TMS-evoked EEG responses masked by muscle artifacts. NeuroImage, 139:157-166. doi:10.1016/j.neuroimage.2016.05.028