The bispectral index (BIS) monitor is widely used to assess depth of anesthesia, particularly
during total intravenous anesthesia (TIVA). We report a limitation observed during
intraoperative D-wave monitoring for intradural extramedullary (IDEM) tumor excision
under TIVA without neuromuscular blockade.
Between January 2024 and May 2025, 14 IDEM tumor excisions were performed with intraoperative
D-wave monitoring using the NIM-Eclipse E-4 system (Medtronic, United States) with
the electrode placed one level caudal to dural opening ([Supplementary Fig. S1], available in online version only). Continuous single-pulse transcranial stimulation
at 1 Hz was employed. In every case, stimulation produced a high-voltage artifact
at the same frequency in the BIS-processed electroencephalographic (EEG) waveform
([Fig. 1]). This was visually disruptive and consistently prevented BIS computation. Since
TIVA is often combined with D-wave and motor evoked potential monitoring, reliable
EEG-based depth assessment becomes essential.[1] Artifact contamination thus rendered BIS monitoring unreliable at precisely the
period when it was most needed.
Fig. 1 BIS-processed EEG waveform displaying high-voltage artifacts (indicated by arrow)
corresponding to 1 Hz D-wave stimulation during intraoperative neuromonitoring. The
artifact results in a BIS Signal Quality Index (SQI) of 0 and elevated electromyographic
(EMG) activity (75%), leading to failure of BIS value computation. EEG, electroencephalography.
To mitigate this issue, we increased the stimulation frequency to 2 Hz in a 55-year-old
woman undergoing IDEM excision at D9 to D12 (SDC 1). This was based on the assumption
that BIS and NIM Eclipse filters (typically 0.5–1 Hz) would attenuate artifacts below
this cutoff, preserving higher frequency EEG for spectral edge frequency computation.[2] Contrary to expectation, the BIS monitor continued to show artifacts, now at 2 Hz,
while failing to compute the index, and the NIM Eclipse EEG trace was entirely lost.
This adjustment was repeated in subsequent cases, but artifacts persisted. We also
attempted higher high-pass filter settings (≥2 Hz) on the NIM Eclipse in one case;
the artifact persisted, confirming that pulse amplitude rather than frequency alone
was the major contributor.
The mechanism lies in the nature of D-wave stimulation. Although repeated at 1 Hz,
each pulse is a short, high-voltage impulse with wide frequency content. By Fourier
theory, such impulses contain harmonics across the EEG spectrum, and their large amplitude
saturates multiple frequency bands simultaneously.[2] As a result, BIS input channels are overwhelmed and the algorithm, despite using
spectral separation methods, rejects the signal once artifact thresholds are exceeded.
This explains why neither filter adjustments nor frequency shifts restored reliable
computation.
It is well known that electrical stimulation can contaminate EEG.[3] For example, electrocautery generates continuous high-frequency interference that
BIS may misinterpret as beta/gamma EEG, producing spuriously high BIS values even
during deep anesthesia. By contrast, D-wave stimulation produces discrete, high-amplitude
broadband impulses that overwhelm the system and result in complete loss of BIS computation.
Both represent limitations of processed EEG monitoring, but the mechanism and consequences
differ: electrocautery risks anesthetic overtitration through falsely high BIS values,[4] whereas D-wave monitoring deprives the clinician of depth information altogether.
Moreover, electrocautery artifacts are typically intermittent, allowing BIS monitoring
to resume once cautery is suspended, while D-wave stimulation is delivered continuously
for the duration of decompression, leaving no interruption-free window for depth assessment.
Vaithialingam et al recently reported similar artifacts in BIS traces during D-wave
monitoring and suggested using these as surrogate confirmation of stimulus delivery.[5] While this may reassure neurophysiologists, it compromises the anesthesiologist's
ability to assess depth—an arguably more critical function during long spinal tumor
resections. Dedicated neuromonitoring platforms such as NIM Eclipse already confirm
stimulus delivery, making BIS artifact recognition redundant.
While artifacts from electrocautery and other electrical devices are well recognized
and frequently reported, interference from D-wave stimulation remains highly underreported
despite being consistently encountered in practice. Our observation underscores the
importance of reporting such artifacts, not only to improve awareness among clinicians
but also to drive advancements in EEG-based depth monitoring algorithms and hardware
design.[6] Highlighting these limitations is essential if we are to refine neuromonitoring
strategies and ensure reliable anesthesia depth assessment during complex spinal procedures.
