CLINICAL STUDIES
1064 | VOLUME 66 | NUMBER 6 | JUNE 2010 www.neurosurgery-online.com
Julian Prell, MD
Department of Neurosurgery,
University of Halle,
Halle, Germany
Jens Rachinger, MD
Department of Neurosurgery,
University of Halle,
Halle, Germany
Christian Scheller, MD
Department of Neurosurgery,
University of Halle,
Halle, Germany
Alex Alfieri, MD
Department of Neurosurgery,
University of Halle,
Halle, Germany
Christian Strauss, MD
Department of Neurosurgery,
University of Halle,
Halle, Germany
Stefan Rampp, MD
Department of Neurosurgery,
University of Halle,
Halle, Germany
Reprint requests:
Julian Prell, MD,
Department of Neurosurgery,
University of Halle,
Ernst-Grube-Str 40,
06097 Halle, Germany.
E-mail:
julian.prell@medizin.uni-halle.de
Received, February 3, 2009.
Accepted, December 5, 2009.
Copyright © 2010 by the
Congress of Neurological Surgeons
R
elevant facial nerve paresis is seen frequently
in the immediate postoperative period,
especially in patients with large vestibular
schwannomas. The cosmetic and social implica-
tions are of critical importance for the patients.
1-4
Consequently, the prevention of facial paresis is
one of the main goals in vestibular schwannoma
surgery. Although anatomic integrity of the nerve
is preserved in most cases,
2,5-7
functional impair-
ment ranging from slight postoperative paresis
to total paralysis is still seen in up to 70% of all
patients.
1,3,8-11
As a consequence, neurosurgeons
have always desired reliable methods capable of
foreseeing paresis and ultimately its extent. A
multitude of techniques for intraoperative facial
nerve monitoring have been developed to reach
this goal.
12-21
Apart from methods that are based
on stimulation techniques, providing informa-
tion about the functional status of the facial nerve
at distinct and discrete points of time during sur-
gery, the free-running electromyogram (EMG)
as introduced by Prass and Lüders
17
in 1986 has
been the technique of choice for >2 decades.
Visual analysis of all recorded EMG activity
20
has allowed differentiation between various pat-
terns of prolonged, surgery-related activity com-
monly referred to as “neurotonic discharges.”
22
One of these patterns, the A-train, has been
shown to be pathognomonic for postoperative
facial paresis.
2,20,21,23,24
The A-train is a pattern
composed of similar elements with high interpeak
frequency. Additional characteristics are sudden
onset and sudden termination. Its duration varies
between milliseconds and several seconds. The
normal amplitudes of A-trains are comparatively
low, usually presenting with 100 to 200 μV
(Figure 1).
Subsequently, an algorithm for automated
detection and quantification of A-trains by a com-
puter program has been developed. With this
algorithm, all episodes of the facial nerve EMG
ABBREVIATIONS: EMG, electromyogram; HB,
House and Brackmann
A Real-Time Monitoring System for the Facial Nerve
OBJECTIVE: Damage to the facial nerve during surgery in the cerebellopontine angle is indi-
cated by A-trains, a specific electromyogram pattern. These A-trains can be quantified by
the parameter “traintime,” which is reliably correlated with postoperative functional out-
come. The system presented was designed to monitor traintime in real-time.
METHODS: A dedicated hardware and software platform for automated continuous analy-
sis of the intraoperative facial nerve electromyogram was specifically designed. The auto-
matic detection of A-trains is performed by a software algorithm for real-time analysis of
nonstationary biosignals. The system was evaluated in a series of 30 patients operated on
for vestibular schwannoma.
RESULTS: A-trains can be detected and measured automatically by the described method
for real-time analysis. Traintime is monitored continuously via a graphic display and is
shown as an absolute numeric value during the operation. It is an expression of overall,
cumulated length of A-trains in a given channel; a high correlation between traintime as
measured by real-time analysis and functional outcome immediately after the operation
(Spearman correlation coefficient [ρ] = 0.664, P < .001) and in long-term outcome (ρ =
0.631, P < .001) was observed.
CONCLUSION: Automated real-time analysis of the intraoperative facial nerve electromyo-
gram is the first technique capable of reliable continuous real-time monitoring. It can crit-
ically contribute to the estimation of functional outcome during the course of the operative
procedure.
KEY WORDS: A-train, Electromyogram, Facial nerve, Monitoring, Real-time, Vestibular schwannoma
Neurosurgery 66:1064-1073, 2010 DOI: 10.1227/01.NEU.0000369605.79765.3E www.neurosurgery-online.com
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/66/6/1064/2556787 by guest on 12 June 2020