The Human Burst Suppression Electroencephalogram of Deep Hypothermia

Abstract

This repository contains the data and code to reproduce the results from "The Human Burst Suppression Electroencephalogram of Deep Hypothermia". The paper investigates the electroencephalogram (EEG) changes and burst suppression patterns induced by deep hypothermia during cardiac surgery. The repository includes the raw EEG recordings from 11 patients undergoing deep hypothermia, as well as the code used for detection of bursts and suppression, spectral analysis, and other analyses from the published manuscript. The results support a model of cyclical metabolic depletion underlying burst suppression, and suggest local micro-network dropout explains decreasing burst amplitudes at lower temperatures. 

Background

Deep hypothermia induces ‘burst suppression’ (BS), an electroencephalogram pattern with low-voltage ‘suppressions’ alternating with high-voltage ‘bursts’. Current understanding of BS comes mainly from anesthesia studies, while hypothermia-induced BS has received little study. We set out to investigate the electroencephalogram changes induced by cooling the human brain through increasing depths of BS through isoelectricity.

Methods

We recorded scalp electroencephalograms from eleven patients undergoing deep hypothermia during cardiac surgery with complete circulatory arrest, and analyzed these using methods of spectral analysis. Full details are provide in the published manuscript.

Data Description

Table 1 in the supplemental material of the published manuscript summarizes the patient characteristics and drugs administered during general anesthesia. Induction of anesthesia was with propofol (n=7), etomidate (n=2), and midazolam (n=2). For intubation muscle relaxation was achieved with a non-depolarizing muscle relaxant (atracurium, cisatracurium, succinylcholine, or vecuronium). For intubation muscle relaxation was achieved with a non-depolarizing muscle relaxant (atracurium, cisatracurium, succinylcholine, or vecuronium). General anesthesia outside the period of deep hypothermic circulatory arrest (DHCA) was maintained with either isoflurane (0.5-1.4% measured at end expiration, n=10) and intravenous narcotics (fentanyl, hydromorphone, or morphine), or in a single case with propofol 150mcg/kg/min) and sulfentanil (0.5 mcg/kg/hr). Throughout DHCA when patients were on bypass, the sole anesthetic was either isoflurane, administered continuously directly into the blood at a concentration of 1% (n=10) or propofol (150mcg/kg/hr, n=1).

EEG Recording and Anesthesia 
All patients were anesthetized with midazolam, fentanyl, isoflurane, and a non-depolarizing muscle relaxant. Roller pump cardiopulmonary bypass (CPB) was instituted using standard venous cannulation (bicaval or two stage) and arterial cannulation of either the left femoral artery, right axillary artery or aortic arch connected via heparin-coated tubing, membrane oxygenator (Medtronic, Minneapolis MN) and open hard-shell venous reservoir. The left ventricle was usually vented through the right superior pulmonary vein. Heparin was administered to maintain the activated clotting times (ACT) above 450 seconds during CPB. Nasopharyngeal, bladder, pulmonary artery, arterial inflow, and venous outflow temperatures were continuously monitored. 

Patients were cooled according to a standardized protocol for a minimum of 30 minutes with a maximum inflow/outflow temperature gradient of 10° C and inflow temperatures were not allowed to fall below 15° C. FiO2 was maintained at 1 (100%) during cooling and warming. Alpha-stat blood gas management was used for temperatures below 28° C with CO2 gas added to ventilation gas (approx. 3-5%) to achieve a pCO2 (at temp) = 40 mmHg and pH-stat was used above that blood temperature. Minimum hematocrits were between 25% and 30% depending on patient factors and surgeon’s preferences.

Deep hypothermic circulatory arrest (DHCA) was not initiated until a nasopharyngeal temperature of 18° C was achieved and maintained for 30 minutes and the EEG was isoelectric, as determined by visual analysis. At the time of circulatory arrest the venous line was clamped after an exsanguination of about one liter. Retrograde cerebral perfusion was not generally used. When used, antegrade cerebral perfusion was maintained at 5-15 ml/kg/min. Throughout DHCA, the sole anesthetic administered was isoflurane, which was administered continuously delivered directly into the blood via the bypass circuit at a concentration of 1%. 

At the end circulatory arrest patients were reperfused for 5 minutes at 18-20° C before rewarming. For rewarming the inflow/outflow gradient was again maintained at less than 10° C with temperature maximums of >37° C, >36° C and >35° C for blood, nasopharyngeal and bladder respectively.

Usage Notes

Code to reproduce the figures is provided through the AWS Open Data Sponsorship Program. Code is in the project GitHub repository. To recreate the figures starting from the source data, run the following pieces of code in the following order:

Code: a_Step1_PrepareDataCombined_MakeFig1.m (formerly: a_Step12_PrepareDataCombined.m)
Input: h1_data.mat,... (EEG, temperature,...)
Output: CURATEDDATA1.mat,...; SPECTRA1.mat,...; INDIVBURSTDATA1.mat,...; FigEEG_BSR_Temp_1,... (Fig 1-3. a-e)

Code: a_Step2_Fig1_fg.m
Input: calls: fcnIndividualBurstsInThrees; INDIVBURSTDATA1,...
Output: FigBurstExs_Case1, ... (Fig 1-3, f-g)

Code: fcnIndividualBurstsInThrees(fileNo) (this is a function)
Input: INDIVBURSTDATA1.mat,...
Output: FigBurstExs_Case1, ... (Fig 1-3, f-g)

Code: a_Step2b_Fig1_h.m (formerly: a_FIG3_BurstColorPlots_C_Plot.m)
Input: BurstSpectraTempIndiv_1,...,14, BurstSpectraTempAll
Output: Fig3_ColorPlots.png (Fig 1-3, h)

Code: a_Step3b_Fig4.m (formerly: a_FIG8A_BSPvsTemp.m)
Input: CURATEDDATA1.mat,...
Output: BSPTEMP.mat

Code: a_Step3b_Fig4.m (formerly: Fig8B_BSPvsTemp.m)
Input: BSPTEMP.mat
Output: FigBSPvsTemp.png, FigBSPIQR.png (Fig 4, a-b)

Code: a_Step4_Fig5.m (formerly: a_FIG2B_BurstAmplitudeDistributions_CDFs.m)
Input: SPECTRA1,...
Output: Fig2_LengthAmplitudeCDFs.png (Fig 5) 

Code: a_Step5a_SurivalCurves_Fig6.m (formerly: a_Step20_GetBurstSuppLengthsAndTemps.m)
Input: SPECTRA1,...; CURATEDDATA1,...
Output: SurvivalData 

Code: a_Step5b_SurivalCurves_Fig6.m (formerly: a_Step20C_FindSurvivalCurvesD_RSM.m)
Input: SurvivalData
Output: FigCDFsColorBar.png (Fig 6,a-d)

Code: a_Step6a.m (formerly: a_Step15_MakeDataForWarmColdComparisonGROUP.m)
Input: CURATEDDATA1,...
Output: WarmCold1,...

Code: a_Step6b.m (formerly: a_Step16_MakeDataForLongVsShort.m)
Input: CURATEDDATA1,...
Output: ShortLong1,...

Code: a_Step6c.m (formerly: a_Step16_MakeDataForLongVsShortIndividuals.m)
Input: CURATEDDATA1,...
Output: ShortLong1.mat,...

Code: a_Step7_FIG7_CompareSpectraHotCold.m
Input: WarmCold1,...
Output: Fig_WarmColdSpectra1.png,...

Code: a_Step15_BSRvsTEMP_RampilWay_v2.m
Input: cs1_bursts.mat,...,cs_14_bursts.mat; case8ind.mat (these seem to be manually made)
Output: PlotDataForCaseNo1.mat, ..., 

Ethics

Standard Protocol Approvals, Registrations, and Patient Consents: This retrospective study was conducted under a protocol approved by The Massachusetts General Hospital IRB. Informed consent was not required.
Patient Population. All data are de-identified. 

Conflicts of Interest

None