AUTHOR=Al-Sadek Tamara , Wadhwa Aryan , Wadhwa Millen , Warren Aaron E. L. , Rolston John D. 

TITLE=Methodologies to detect cortico-cortical evoked potentials: a systematic review

JOURNAL=Frontiers in Human Neuroscience

VOLUME=Volume 19 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2025.1636115

DOI=10.3389/fnhum.2025.1636115

ISSN=1662-5161

ABSTRACT=IntroductionCortico-cortical evoked potentials (CCEPs) are electrophysiological responses elicited by direct electrical stimulation of one cortical region and recorded from another, providing insights into functional connectivity and communication pathways between brain areas. However, no consistent standard for defining and measuring CCEPs currently exists.MethodsWe conducted a systematic review of the CCEP literature on detection methods to evaluate commonalities and gaps in methodology. Extracted data included demographics, disease, recording type, montage, recording system, stimulation amplitude and frequency, time window used for epoching around stimulus onset, open access availability, and detection approach.ResultsOf 187 studies undergoing full-text review, over half lacked a description of the CCEP detection method. Specifically, 9.1% utilized visual identification, whereas 49.74% did not explicitly state the method. The remaining 72 studies represented 3,424 patients, of whom 58.3% had sEEG electrodes and most had epilepsy. The most common detection method was threshold-based (68.1%), followed by statistical testing (16.7%) to determine whether CCEPs differed significantly from baseline, data-driven methods (4.1%) that quantify responses after learning from data, and frequency-based approaches (4.1%). Bipolar (48.6%) and single-electrode referential montages (18.1%) were most frequently employed.DiscussionCurrent CCEP detection methods lack consensus, with many studies omitting methodological details and relying heavily on threshold-based techniques that assume fixed response shapes. Future research should encourage the use of data-driven approaches, which learn directly from data, offer more robust alternatives, and improve quantification in both clinical and research contexts.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD42024568261, identifier CRD42024568261.