Medical Student Boston University School of Medicine Boston, MA, US
Introduction: Cortico-cortical evoked Potentials (CCEPs) are an important tool in mapping brain connectivity in clinical and research settings. Accurate CCEP recordings provide the ability to define proper epileptic and tumor resection margins without unnecessarily sacrificing eloquent brain areas. However, the detection of CCEP responses lacks a defined systematic approach, impacting the reliability and reproducibility of results in clinical settings. Thus, we aimed to systematically review the breadth of CCEP detection methodology in the literature.
Methods: Preferred reporting items for systematic reviews and meta-analysis guidelines (PRISMA) were utilized in this work. We extracted demographic data, disease type, recording type (stereoelectroencephalography or electrocorticography), montage, CCEP recording system, stimulation intensity, stimulation frequency, epoch limits, whether or not the code was open source, and the CCEP detection method used.
Results: After screening 1990 studies, a total of 57 studies met inclusion criteria, representing 3199 patients, with 57.9% having sEEG electrodes implanted. Most of the patient population had epilepsy (n=51, 89.5%), with some having tumors (n=2, 3.5%) or both (n=3, 5.3%). The most common detection method was threshold or peak-based (67.2%), followed by statistical testing (17.5%), data-driven (5.3%), and frequency-based approaches (5.3%). The re-referencing montages utilized were mainly bipolar (47.4%) followed by referential (17.5%) and adjusted common average reference (7.0%) among others.
Conclusion : This review demonstrates the lack of standardization in detection methods available for CCEPs, underscoring the need for automated, consistent, and reliable methodologies. Choosing more data-driven methods like Canonical Response Parametrization (CRP) offers a systematic approach to quantifying evoked responses, leading to greater specificity in the treatment and targeting of various diseases.
.jpg)
