Abstract Details

Title Characterization and Validation of Somatosensory Evoked Potential Morphology during Carotid Endarterectomy to Detect Stroke

Topic Cerebrovascular Disease and Interventional Neurology

Presentation(s) P1 - Poster Session 1 (5:30 PM-6:30 PM)

Poster/Presentation
Number 3-064

Objective Improve stroke detection rates during carotid endarterectomy with a novel method to characterize somatosensory evoked potentials (SSEPs) during intraoperative neuromonitoring (IONM).

Background IONM changes in peak-to-peak voltage and latency can be used to detect periprocedural neurological deficits such as stroke after plaque removal in carotid endarterectomy. However, these manual marker measurements are subjective to bias and could potentially fail to take into account atypical SSEPs, for example traces without discernible peaks. In contrast, phase state analysis allows for automatic characterization of changes in SSEP morphology.

Design/Methods Retrospective study of 116 patients who underwent a carotid endarterectomy procedure. IONM was performed with left and right median or ulnar nerve stimulation and P4/Fz and P3/Fz SSEP measurements were recorded. SSEPs were collected at baseline, before and after clamping, during plaque removal and after clamp removal (averaged >2 trials at every stage). SSEPs were plotted in phase space (slope vs. voltage value) and the Phase Space Area (PSA) was calculated. A 50% change in PSA was defined as a significant change. SSEPs were also individually reviewed and peak-to-peak and latencies were manually obtained as a control.

Results 25 patients had a stroke during the CEA procedure. PSA analysis was subjected to a 10-fold cross validation scheme and yielded a sensitivity of 68% (95% CI:47-85), specificity of 81% (95% CI:62-93), PPV of 74% (95% CI:51-89) and NPP 77% (95% CI:57-89) in detecting stroke. Contrastingly, peak-to-peak (<50% from baseline) or latency (>10% from baseline) changes yielded a sensitivity of 35% (95% CI:24-45), specificity of 88% (95% CI:79-93), PPV of 35% (95% CI:14-61) and NPP 80% (95% CI:71-88).

Conclusions Phase state analysis was validated with traditional SSEP measurements. Moreover, phase state analysis was able to detect periprocedural stroke occurrences after cross clamping in a CEA procedure with comparable accuracy to traditional methods.

Authors/Disclosures
Ahmed Jorge PRESENTER	No disclosure on file
Donald Crammond	No disclosure on file
Jeffery Balzer	No disclosure on file
Katherine M. Anetakis, MD (UPMC)	An immediate family member of Dr. Anetakis has received personal compensation in the range of $5,000-$9,999 for serving as an Expert Witness for The Rosenthal Firm.
Parthasarathy Thirumala, MD, FAAN (University of Pittsburgh Medical Center)	The institution of Dr. Thirumala has received research support from University of Pittsburgh.

��ɫ��

��ɫ��