Send to

Choose Destination
J Clin Neurophysiol. 2019 Mar 14. doi: 10.1097/WNP.0000000000000570. [Epub ahead of print]

Crossover Phenomena in Motor Evoked Potentials During Intraoperative Neurophysiological Monitoring of Cranial Surgeries.

Author information

Surgical Neurophysiology Program, Department of Neurology, University of Southern California, Los Angeles, California, U.S.A.



Transcranial motor evoked potentials (TcMEPs) are used to assess the corticospinal tract during surgery. Transcranial motor evoked potentials are elicited by preferentially activating the anode over the target cortex. Crossover occurs when stimulation also induces activation of ipsilateral motor evoked responses. These responses are believed to be generated by activation of corticospinal tract on more caudal neural structures. The presence of cross activation poses a problem in craniotomy surgeries because activation of neural structures occurs distal to the area of interest leading to false negatives. Eliminating crossover may lead to activation of the motor pathway proximal to the surgical site, thus potentially reducing false-negative responses. There are no data on how often crossover signals occur or the conditions in which they take place. This study examines the frequency of crossover, the surgical procedures in which they occur, and their stimulation parameters.


We reviewed all the TcMEP data files for intracranial procedures performed in 2016 at Keck Hospital of USC. We recorded demographic information about the surgical side, lobe, diagnosis, age, and sex. Only baseline TcMEPs were analyzed. Crossover responses were deemed present if recorded amplitudes were greater than 25μv on the ipsilateral side. We evaluated the rate of crossover presence, the lowest voltage associated with crossover, the highest voltage without crossover, if crossover resolved, and the last muscles to remain present when crossover is eliminated. Transcranial motor evoked potentials were divided into four groups. Group A: crossover present and was not resolved, group B1: crossover present but resolved with desired signals, group B2: no crossover seen with desired signals in both limbs, and group C: crossover resolved with loss of signals in either limb. The Difference between lowest amplitude with crossover and highest amplitude without crossover was obtained for each patient, and the mean of this difference was calculated using paired t-test.


We analyzed 186 TcMEPs. Forty-four TcMEPs were in group A, 52 in B1, 68 in B2, and 22 TcMEPs were in group C. Of total crossovers (118), 63% resolved at baseline, whereas 37% did not resolve. The mean difference between minimum value with crossover and maximum value without crossover was 50 V (P < 0.0001). In five TcMEPs, this difference was 0 and the median was 250 V. There was no significant difference between surgical site, stimulation side, pathology, or sex between crossover (A) and noncrossover (B + C) groups. There was a significant association found between age group ≤50 years versus >50 years and being in crossover versus noncrossover groups (P = 0.01). For 95% of the cases in group C, the last muscles to stay were hand muscles.


Transcranial motor evoked potential crossover may pose a problem during surgeries leading to false-negative results. Crossover is a frequent phenomenon that should not be overlooked. Stimulation intensity is the main factor contributing to the reduction of crossover. Crossover can be reduced in most TcMEPs performed (63%) leading to adequate monitoring in 76% of TcMEPs. Despite best efforts, there are still one quarter (24%) of TcMEPs where crossover cannot be eliminated. Newer strategies should be sought to reduce crossover. Teams should focus their efforts on reducing crossover of TcMEPs to make monitoring of intracranial surgeries more reliable.

Supplemental Content

Full text links

Icon for Wolters Kluwer
Loading ...
Support Center