Dhritiman Chakrabarti

ECG contamination of EEG signals: effect on entropy

Entropy™ is a proprietary algorithm which uses spectral entropy analysis of electroencephalographic (EEG) signals to produce indices which are used as a measure of depth of hypnosis. We describe a report of electrocardiographic (ECG) contamination of EEG signals leading to fluctuating erroneous Entropy values. An explanation is provided for mechanism behind this observation by describing the spread of ECG signals in head and neck and its influence on EEG/Entropy by correlating the observation with the published Entropy algorithm. While the Entropy algorithm has been well conceived, there are still instances in which it can produce erroneous values. Such erroneous values and their cause may be identified by close scrutiny of the EEG waveform if Entropy values seem out of sync with that expected at given anaesthetic levels.

Endovascular coil detachment causing EMG artefact in BIS: a mechanistic exploration

Abstract Deployment of endovascular coils used in interventional neuroradiology commonly involves electrolytic detachment of the coil from the pusher catheter. This report describes a case of artefactual increase in electromyography (EMG) values of bispectral index (BIS) monitor during coil detachment. An explanation of this event is provided connecting mechanism of coil detachment and derivation of EMG values in a BIS monitor. While rising EMG values are thought to arise from frontalis contraction, they may as well be an unrecognized electrical artefact, especially in context of undistorted electroencephalography waveform.

Can Intraoperative Neurological Preconditioning Occur After Intraoperative Hypotensive Episodes

To JNA Readers: Ischemic preconditioning has been proven experimentally, especially in animal studies to improve tolerance of neuronal tissue to subsequent ischemic insults. A Chinese study in subarachnoid hemorrhage patients showed intraoperative short-term controlled temporary clipping of cerebral vessel to be beneficial. It reduces decline of cerebral oxygen tension and pH in patients whose aneurysm structure required temporary clipping of vessel.1 We report a case of internal carotid artery (ICA) giant aneurysm with evidence of intraoperative preconditioning due to periods of spontaneous hypotension related to blood loss. Written consent of the patient was obtained before writing this report. The case was a 62-year-old gentleman, with symptoms of sudden-onset headache consistent with subarachnoid hemorrhage. Contrast-enhanced CT scan of the brain showed a thrombosed right cavernous ICA aneurysm, confirmed by cerebral digital subtraction angiography. Endoscopic transnasal transethmoidal clipping of aneurysm was planned with neuronavigation. Because of requirement of temporary clipping of ICA intraoperatively, the anesthetic plan involved monitoring of bilateral frontotemporal electroencephalography montages FP 1/2 to T3/4 (EEG) and frontal cerebral oximetry (SctO2). EEG and SctO2 probes were applied after induction, positioning the patient and registering the facial landmarks in neuronavigation. The baseline EEG showed low amplitude a and d activity with spectral edge frequency of 7 to 8Hz on both sides with baseline SctO2 of 68% to 72% bilaterally and systolic blood pressure (SBP) of 140mm Hg. After dissecting through the bone, ICA was visualized with aneurysm and the ICA was temporarily clipped. The ipsilateral SctO2 reduced to 64% (contralateral 71%) within 10 minutes, whereas the EEG did not show any amplitude or frequency changes. Thus, collaterals were presumed to be adequate and the surgeon was informed. During aneurysm dissection, aneurysm ruptured with acute blood loss of approximately 0.5L, and SBP decreased to 100mm Hg. At this pressure level, the ipsilateral SctO2 was noticed to reduce to 55% with contralateral maintaining 70% and spectral edge frequency decreasing slightly to 6 to 7Hz bilaterally. Rapid fluid infusion and blood transfusion improved SBP to 120mm Hg and SctO2 was seen to improve to prehypotension levels of 64. We surmised the SBP threshold for ipsilateral hypoperfusion to be 100mm Hg. Blood loss continued at a rapid rate, and 15 minutes later, the SBP decreased to 90mm Hg despite replacement, and the SctO2 declined below 90mm Hg. As noted previously, it improved when blood pressure was brought up to 120mm Hg. Thirty minutes later SBP reduced to 80mm Hg temporarily, but the SctO2 was seen to stay stable at 62% to 64%. After multiple clipping attempts, bleeding

Intraoperative aberrant bispectral index values due to facial nerve monitoring

Bispectral index is an accepted depth of anaesthesia monitor for guiding intraoperative hypnotic agent administration. Frontalis EMG displayed on BIS monitor may increase due to twitching of frontalis muscle. EMG increases are also known to cause artefactual increases in BIS values. We report a case of artefactual increase of EMG and subsequently BIS values, due to electrical artefact from cranial nerve stimulator being used to identify the facial nerve. An explanation of the effect of stimulator signal on BIS EMG and BIS values has been provided.

Adverse Hemodynamic Event Due to Floseal Hemostatic Matrix Application.

after 20 minutes was found to be 3.8% of the baseline. We theorized that the remaining ICG caused small decreases of SctO2 to be hidden and thus, pseudonormal values were displayed. Although further studies are required to quantify the effect of the remaining ICG on SctO2 to prove clinical significance of this finding, it would be wise to use multimodality monitoring to monitor functional aspect of neurons (EEG, evoked potentials) along with NIRS-based SctO2 when ICG administration is planned.

Brainstem Contusion: A Fallacy of GCS-BIS Synchrony.

To JNA Readers: The correlation between the Glasgow Coma Scale (GCS) and the Bispectral Index (BIS) in traumatic brain injury patients has been demonstrated in multiple studies. The BIS score has also been shown to be a predictor of the outcome in this group of patients. However, the BIS value is only representative of the cortical activity of the frontal lobe over which the electrode has been placed. We present here a case of traumatic acute subdural hematoma (SDH) with midbrain contusion and poor GCS, but apparently normal cortical electrical activity as evidenced by BIS monitoring. The case was of a 46-year-old male patient with traumatic right fronto-temporo-parietal acute SDH (volumeE45mL, midline shift 8.9mm on noncontrast computed tomographic [CT] scan, approximately 9h after the injury). The patient’s admission GCS was E1V1M2, and the decision was made for emergent evacuation of SDH. The BIS electrode was placed before induction on the opposite side as described by Nelson et al3 to prevent intraoperative disruption of readings due to blood soakage. Unexpectedly, the baseline BIS of the patient was 92, with a spectral edge frequency of 25Hz. We had expected a lower BIS value due to the low GCS score. Induction of anesthesia was performed with injection thiopentone and maintained with sevoflurane with a target BIS between 40 and 60. About 0.7 to 0.8MAC of sevoflurane was required for maintenance. The intraoperative course was uneventful. Postoperative anesthetic withdrawal led to an increase in the BIS to preoperative levels, and the GCS was status quo. Because of a discrepancy between the BIS values and the GCS, the EEG waveform obtained from the BIS monitor was evaluated during and after the surgery, but spikes suggestive of seizure activity (which may artifactually elevate the BIS value) were not noted. The patient maintained a normal pattern of spontaneous respiration and hemodynamics in the postoperative period. The CT scan was reviewed again, and specks of hemorrhage were noted on the posterolateral aspect of the midbrain bilaterally with adjacent parenchymal hypodensity (Fig. 1A). Brainstem contusion was thus surmized as the cause of low GCS in a patient with apparently intact cortical electrical activity. The postoperative day 1 CT scan showed the contusions on the midbrain more clearly (Fig. 1B). Multiple applications of BIS since its introduction have the common theme of analyzing the superficial cortical neuronal activity. The apparent correlation of BIS and GCS in traumatic brain injury is based on the premise of fronto-temporal neuronal dysfunction. Our case illustrates that it is possible for a patient to have apparently normal cortical neuronal activity and a decreased level of consciousness due to damage to subcortical structures. On scrutinizing the CT scan, the brainstem injury was localized to the tegmentum, which houses the reticular formation and the red nucleus, damage to which may result in the impairment of consciousness and decerebrate posturing. A study by Paul et al1 on 29 mild to moderate head injury patients revealed a strong positive correlation between BIS values and GCS (r=0.67; P<0.001). However, they noted that the excessive scatter of the BIS values prevented the prediction of GCS from a single BIS value. In our experience, on the basis of unpublished data, we noted that headinjured patients with low GCS scores tend to have low baseline BIS values FIGURE 1. A, The preoperative computed tomographic (CT) scan showing specks of hemorrhage on the midbrain surface (black arrows). B, The postoperative day 1 CT scan shows well-defined midbrain contusions (white arrows).

Fluctuating entropy values during frontal craniotomy

Electroencephalogram (EEG) entropy is a widely used monitor of depth of anaesthesia. We present a series of two cases of frontal craniotomies in which entropy monitoring was unreliable due to wide fluctuations. Also, a small experiment was conducted to identify the effect of individual electrode displacement on entropy values. We conclude that caution should be exercised in interpreting frontal EEG derived indices of the depth of anaesthesia in frontal craniotomies.

Intraoperative electroencephalography changes in unilateral moyamoya phenomenon

Address for correspondence: Dr. Dhritiman Chakrabarti, Department of Neuroanaesthesia, National Institute of Mental Health and Neuro Sciences, 3rd Floor, Faculty Block, Hosur Road, Bengaluru ‐ 560 029, Karnataka, India. E‐mail: dhritiman.ch@gmail.com EEG was placed for ischaemia monitoring, due to the possibility of ICA handling. Due to the agitated state of child, EEG electrode was placed after induction of anaesthesia. EEG monitoring was instituted with P4‐T6 and P3‐T5 bipolar montages. Electrodes were placed, and secured carefully using gauze and waterproof adhesive tape so as to avoid hindering the surgical field and to prevent blood soakage. Induction with thiopentone, fentanyl and rocuronium was, followed by maintenance with sevoflurane and nitrous oxide 50% targeted to maintain 1.2 age‐adjusted minimum alveolar concentration (MAC).

Longer times to reanesthetization with longer anesthetic duration

To the Editor Leeson et al. 1 address an important clinical problem. The authors simulated 1, 2, 4, and 6 hours of inhalation of isoflurane, sevoflurane, and desflurane at concentrations of 0.75, 1.0, and 1.5 minimum alveolar concentration (MAC). After the simulated inhalation, they reduced the inspired anesthetic concentration to zero and ventilated the simulated subjects with 4.0 L/min until the anesthetic concentration in the vessel-rich group (i.e., brain) decreased to 0.33 MAC, simulating awakening. They then simulated hypoventilation by reducing the minute ventilation to 0 and 0.1 L/min and looked at the time for the anesthetic concentration in the vessel-rich group to exceed 0.5 MAC. They analyzed the times to reanesthetization as a function of minute ventilation, duration of To the Editor We read with interest the meta-analysis by De Oliveira et al.1 and subsequent Letter to the Editor by Groudine,2 which suggest that the analgesic effect of the transversus abdominis plane (TAP) block may be related to both the local neural blockade and a systemic effect of absorbed local anesthetics. We believe this is likely, as only 3 to 3.75 nerves are typically blocked after classic TAP block, and only 5.5 nerves are blocked after subcostal TAP block.3 The mean number of blocked nerves corresponds to the mean number of dermatomes blocked.4–6 Børglum et al.6,7 reported that 7 nerves and dermatomes could be blocked with a combined classic and subcostal technique but with an overall higher efficacy rate in the lower dermatomes.5–7 Not only is block extent limited but there is also a great variability between techniques and in the spread of local anesthetic between muscular fascias.8 Only a few studies report dermatomal assessment as a measure of block effectiveness. Most studies report analgesic consumption. Moreover, dermatomal anesthesia has not been correlated with analgesic consumption. Latzke et al.8 found conflicting results about local anesthetic concentrations in the injected site and noninjected sites as underscored by Groudine.2 While they provided support for neural blockade mechanism, they did not assess dermatomal anesthesia in the injection site and noninjected sites, and their study cannot support a systemic effect of TAP block. Pharmacokinetic data may suggest a systemic mechanism. The terminal half-life of local anesthetics after injection into the abdominal wall is longer than that after IV administration,6 possibly because the terminal half-life is determined by slow absorption from the tissue depot and flip-flop pharmacokinetics. This may explain the longlasting plasma levels observed in patients. Pettersson et al.9 reported a high variability in peak concentration times and terminal half-life times, suggesting a marked variation in absorption rates after abdominal blocks. These investigators reported that patients with a longer terminal half-life of ropivacaine asked for their first postoperative analgesic later than patients with a shorter terminal half-life. Because both local and systemic effects of local anesthetics may contribute to the analgesic response, we believe that pharmacokinetic findings should be correlated to sensory outcomes and analgesic consumption to understand the role of systemically absorbed drugs.