Accepted Manuscript Tracking EEG changes during the exposure to hyperbaric oxygen Lucio Pastena, Emanuela Formaggio, Silvia Francesca Storti, Fabio Faralli, Massimo Melucci, Riccardo Gagliardi, Lucio Ricciardi, Giovanni Ruffino PII: DOI: Reference:
S1388-2457(14)00288-0 http://dx.doi.org/10.1016/j.clinph.2014.05.013 CLINPH 2007110
To appear in:
Clinical Neurophysiology
Accepted Date:
5 May 2014
Please cite this article as: Pastena, L., Formaggio, E., Storti, S.F., Faralli, F., Melucci, M., Gagliardi, R., Ricciardi, L., Ruffino, G., Tracking EEG changes during the exposure to hyperbaric oxygen, Clinical Neurophysiology (2014), doi: http://dx.doi.org/10.1016/j.clinph.2014.05.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Tracking EEG changes during the exposure to hyperbaric oxygen
Lucio Pastenaa, Emanuela Formaggiob,§, Silvia Francesca Stortic,§,*, Fabio Farallid, Massimo Meluccid, Riccardo Gagliardid, Lucio Ricciardid, Giovanni Ruffinod
a
Department of Neurological Sciences, University of Rome, La Sapienza, Rome, Italy
b
Department of Neurophysiology, Foundation IRCCS San Camillo Hospital, Venice, Italy
c
Department of Neurological and Movement Sciences, University of Verona, Verona, Italy
d
Italian Navy Medical Service Comsubin Varignano, Le Grazie (La Spezia), Italy
§
These authors equally contributed to the study.
1
* Corresponding author: Silvia Francesca Storti Dipartimento di Scienze Neurologiche e del Movimento Policlinico G.B. Rossi – P.le L.A. Scuro 10 – 37134 Verona, Italy Tel: +39 045-8126394 Fax: +39045-8027276 E-mail: [email protected] Abstract Objective: The aim was to investigate and define possible alterations in cerebral activity during prolonged hyperbaric oxygen exposure and decompression as compared to baseline activity. Methods: Thirty-two channel electroencephalography (EEG) was recorded with a Bluetooth EEG system in 11 subjects. A 20-minute EEG recording was carried out under three different conditions: breathing air inside a hyperbaric chamber at sea level; breathing oxygen at a simulated depth of 18 msw; breathing air at sea level after decompression. Relative EEG power was estimated in frequency ranges. Results: During oxygen breathing, brain activity showed an early fast delta decrease in the posterior regions, with a synchronous and significant increase in alpha in the same regions. After decompression, the delta
2
relative power decrease was uniformly distributed over the cerebral cortex until minute 8, and the alpha relative power was maximal in the posterior regions during the first 2 minutes. Conclusions: These results may be relevant for establishing a reference point in future studies on oxygen-sensitive subjects who reported problems during oxygen diving. Significance: Significant changes in EEG relative power suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygen-toxicity diving-related problems".
Keywords: EEG brain mapping; oxygen toxicity; brain activity; EEG frequency analysis.
Highlights -
We investigated EEG changes in professional divers after oxygen breathing at a simulated depth of 18 msw and after decompression breathing air at sea level.
-
A significant decrease in delta and an increase in alpha relative power in the posterior regions were observed during oxygen breathing.
-
The results suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygentoxicity diving-related problems".
1. Introduction The use of oxygen mixture and closed circuit apparatus in diving at various depths augments the risk of central nervous system oxygen toxicity (CNS O2T). Oxygen poisoning can manifest itself with symptoms ranging from nausea, dizziness, hearing and visual disturbance, amnesia, vertigo, irritability, localized
3
muscular twitching, to loss of consciousness, tonic clonic seizures, and EEG alterations such as an increase in theta waves, a decrease in alpha waves, microsleep and a decline in psychometric performance. Prolonged exposure to hyperbaric oxygen (HBO) will result in permanent paralysis and death (Donald 1947). To prevent this syndrome, prompt recognition of the earliest symptoms, as described after dry or humidified HBO exposure, is vital (Donald 1947; Piantadosi et al., 1979; Butler and Thalmann 1986). An overview of work in this area is given by Harabin and Donald (Donald 1947; Harabin 1993). More recently, Arieli et al. reviewed 2527 dive reports to analyze the relationship between symptom onset and its dependence on depth and dive time (Arieli et al., 2006). In some animals, environmental factors, such as cold water and darkness, can increase oxygen sensitivity (Bitterman et al., 1986). During the selection of military divers, an oxygen tolerance test (OTT) can be used to identify subjects with susceptibility to CNS O2T (Butler and Knafelc 1986; Butler and Thalmann 1986). The Italian Navy OTT procedure involves exposing divers to 2.8 atmosphere absolute pressure (ATA) for 15 min in a dry chamber breathing 100% O2 through a face mask. The electroencephalogram (EEG) pattern is known to significantly change during saturation dives (Lemaire and Rostain 1988; Okuda et al., 1988 Pastena et al., 1999), but few researchers have studied EEG modifications in humans during oxygen breathing in a hyperbaric chamber. The main contribution to the study of CNS O2T was made by Visser et al. (1996a,b). Quantitative EEG findings are described for a group of 23 subjects during HBO exposure and for one subject who had a generalized tonic-clonic seizure on exposure to HBO. The subjects were exposed to 100% O2 for 30 min at 2.8 bar. EEG changes were minor and not considered indicative of a HBO effect on the brain in the group of subjects who showed no signs of toxicity. Pre-convulsive EEG changes were detected in the subject who experienced a seizure but they were too insignificant and did not clearly herald clinical signs (Visser et al., 1996a). In the same group of subjects, the blood flow velocity in the right middle cerebral artery was monitored by transcranial Doppler (TCD). The TCD mean velocity
4
decreased during HBO exposure probably due to vasoconstriction of the cerebral resistance vessels, whereas it increased in the subject who experienced a generalized seizure. Importantly, however, no definitive conclusions can be drawn owing to the small number of subjects with toxicity in this study (Visser et al., 1996b). Stronger EEG abnormalities can be found already in the compression phase. For example, during the compression phase of a saturation dive reaching 250 msw and breathing O2 at different pressures (pO2 0.44–0.48 ATA), an increase in theta and beta rhythms in the central regions of the brain were observed during the entire compression period, whereas an increase in delta was noted at a depth of 100 msw (Pastena et al., 1999). In extremely stressful situations, a theta increase could be considered an expression of altered activity in the frontal central areas while maintaining cerebral homeostasis in the presence of remarkable environmental stressors. The increase in delta rhythm could be the epiphenomenon of partial cortical deafferentation in the posterior areas. Rostain and Charpy studied EEG modifications and psychometric performance in healthy subjects during two simulated dives at 500 and 610 m with a helium-oxygen-breathing mixture (pO2 0.4 ATA). The EEG modifications considered as high-pressure nervous syndrome (HPNS) began at 300 m. They were remarkable for an increase in slow activity (particularly theta) in the anterior and middle regions of the scalp, depression of fast activity, and transformation of the waking EEG into one resembling that of stage I sleep. In addition, the sensorimotor performance scores on psychometric testing were lower in all subjects (Rostain and Charpy 1976). Changes in electrical cortical activity have been observed during convulsions induced by an increase in oxygen pressure (Cohn and Gersch 1945). Research into the prevention of CNS O2T has been informed by EEG findings in animals during the preconvulsive period of HBO exposure; however, these studies described only the onset of “irritative” figures (spikes, polyspikes) in many subjects, without any
5
definitive results. Today, EEG brain mapping combined with fast Fourier transform analysis can reveal variations in bioelectrical activities associated with the environmental factor under study, for example, changes in EEG cortical activity during hyperbaric exposure (Zal’tsman 1967), xenon anaesthesia (Morris et al., 1955), saturation dives (Pastena et al., 1999), and nitrogen narcosis (Pastena et al., 2005). Hyperbaric oxygen treatment is indicated for a wide range of illnesses (Neubauer 2000; Murata et al., 2005). Murata et al. evaluated the efficacy of repetitive HBO treatments by monitoring quantitative EEG for symptoms of acute carbon monoxide poisoning and prevention of delayed neuropsychiatric sequelae. Examining the effects of repetitive HBO treatment on quantitative EEG in the patients with CO poisoning, they found that the peak alpha frequency and the relative alpha power significantly increased in the occipital regions after repetitive HBO treatments (Murata et al., 2005). From these findings, the authors hypothesized that monitoring of peak alpha frequency may be an objective indicator for the efficacy of HBO treatments to prevent delayed neuropsychiatric sequelae. Although repetitive HBO therapy is known to be effective for their prevention, the precise mechanisms of its action for CO intoxication are not completely understood (Weaver et al., 2002, Takahashi et al., 1998). Furthermore, hyperbaric oxygenation once played an integral role in the treatment of early stroke, where changes in the EEG pattern observed during treatment (Neubauer 2000) aided in guiding neurorehabilitation. Though widely applied in the clinical setting, the effects of oxygen therapy on healthy subjects are not fully understood (Imlay 2003). Quantitative EEG plays a significant role in EEG-based clinical diagnosis and studies of brain function, revealing new possibilities for clinical application in cognitive and behavioural areas (Babiloni et al., 2011; Storti et al., 2013). The quantification of changes in brain rhythms during oxygen breathing is essential to determine a basic pattern in healthy and trained divers in order to detect alterations in subjects at risk of CNS O2T. EEG quantification in normative data could
6
be used as a reference point for future studies that would serve as a landmark of oxygen-induced brain activity and could be useful in the medical treatment of divers reporting problems. Therefore, the aim of this study was to determine how oxygen, assumed at a constant hyperbaric pressure of 2.8 ATA, affects the bioelectrical activity in human professional divers during the compression phase, and to define the EEG patterns of professional divers during this condition as compared to pre- and post-oxygen breathing. One-hour EEGs were recorded in 11 subjects in order to quantify brain activity modifications under three different conditions: air (baseline at sea level), oxygen breathing (at a simulated depth of 18 msw), and air (after decompression).
2. Methods
2.1 EEG with Bluetooth technology The EEG was recorded inside a hyperbaric chamber using a Holter apparatus equipped with Bluetooth wireless technology. Hyperbaric chambers are equipped with portholes, made of transparent resins (i.e., pressure-resistant Plexiglas) through which Bluetooth radio waves are transmitted outside the chamber. The EEG system, located outside the chamber, should be placed next to the porthole in order to obtain more stable acquisition. The EEG system (Ates Medica Device, Verona, Italy) is a portable recorder connected via Bluetooth wireless transmission to a notebook that visualizes the EEG signal (Fig. 1). EEG recordings were performed using 32 Ag/AgCl electrodes (Electrical Geodesic, Inc., Eugene, OR), which are soaked in a physiological-potassium solution for about 5 min before the beginning of the recording. This solution allows a microclimate to develop between the subject’s scalp and the electrode and keeps the impedances low even during long EEG recordings. The net was adjusted so that the 18, Cz, Oz and pre-auricular points were correctly placed according to the international 10/20 system. By virtue of the net’s geodesic
7
tension structure, all electrodes were evenly distributed over the scalp at approximately the same location in all 11 subjects. The data were recorded against a vertex electrode reference (Cz) at a sampling rate of 250 Hz using the software package Geodesic EEG System on Neurotravel technology (Ates Medica Device and Electrical Geodesic, Inc.). An anti-aliasing hardware low-pass filter with a cut-off frequency of 96 Hz was applied. Electrode impedances were verified and kept below 30 kΩ.
2.2 Subjects and experimental protocol The study sample was 11 healthy male navy divers (mean age 46.2 ±4.9 years). Exclusion criteria were a medical history of respiratory problems, sleep disturbances, smokers, and overweight. All subjects were experienced divers with at least 15 years of experience. All had undergone OTTs and none had ever shown signs of CNS toxicity, neither during an OTT nor during operational diving with 100% O2, nitrox or other mixtures. Four belong to the Operational Divers Group and are personnel specialized in free diving with air to 60 msw, nitrox to 54 msw, heliox to 150 msw, and to 250 msw with a minisubmarine or a special suit. The divers in the moderately experienced group possess a license for diving to 60 msw. The internal ethics committee of University approved the experimental protocol and oral and written informed consent was obtained from the subjects before participating in the study. A simulated depth of 18 msw (2.8 ATA) was selected for the study because exposure at this breathing depth is a routine part of navy diver training courses. The hyperbaric chamber used in the study complied with Italian Navy standards for safety equipment and emergency procedures. All subjects were studied individually. They were accompanied during the dive by a technician who assisted the subject if needed and were monitored by closed circuit television. Arousal was maintained throughout the whole recording session by administering an external acoustic stimulation as soon as one 30-second epoch showed a reduction of more than 50% of the background alpha rhythm, according
8
to American Academy of Sleep Medicine scoring rules (Ibner et al., 2007). The procedure was carried out by an experienced neurophysiologist who was also responsible for detecting any epileptiform phenomena. Visual inspection of EEG signals showed no abnormalities. The divers had complete saturation of blood haemoglobin; nonetheless, the true value of pO2, which requires arterial sampling, cannot be performed inside a hyperbaric chamber. Each recording session lasted 20 min, during which the subject reclined on a cot with eyes closed. A baseline 20-min EEG recording was made at 1 ATA breathing air (AIRpre) in an open chamber. A 2-min compression profile (descent rate 9 m min-1) breathing air was used to reach the oxygen stage at a pressure of 2.8 ATA. At this pressure, the subject breathed pure oxygen via the oronasal mask (O2) and a second 20-min EEG recording was acquired. The atmosphere within the hyperbaric chamber was controlled to maintain the total pressure of 2.8 ATA. After the decompression stage, back on air breathing, the EEG of each subject was recorded for 20 min (AIRpost), discarding the first 2 min (ascent rate 9 m min-1). The overall time of EEG recording was about 1 hour (Fig. 2) The chamber was compressed with air, and a separate breathing circuit was used for the oxygen stage. As the chamber was equipped with to a continuous ventilation system, eventual oxygen leaks from the mask were not thought to influence the atmosphere inside the hyperbaric chamber.
2.3 Data analysis The data were processed in Matlab 7 (MathWorks, Natick, MA) using scripts based on EEGLAB 6.01 (EEGLAB toolbox; http://www.sccn.ucsd.edu/eeglab), as well as a dedicated home-made code created especially for this study. The EEG recordings were band-pass filtered from 1 to 30 Hz with a finite impulse response filter. Visible artifacts in the EEG recordings (i.e., eye movements, cardiac activity, and scalp muscle contraction) were removed using an independent component analysis
9
procedure. Electrode 18 (positioned on the nasion) was rejected because of ocular artifact. The data were processed using an average reference. The EEG data were divided into different time intervals for each condition: one interval for the AIRpre condition (1– 20 min) and seven intervals for the O2 and AIRpost conditions (1, 2, 5, 8, 10–12, 15–17, and 18–20 min). Since these time intervals are sufficient to describe the complete phenomenon, the analysis and the representation of the power maps were limited to these windows. The EEG data for each interval were divided into epochs of 2 s. A fast Fourier transform was applied to non-overlapping epochs, each containing 500 data points for all electrodes and the three experimental conditions, and then averaged across epochs under the same interval condition. The recordings were Hamming-windowed to control for spectral leakage. The power spectra density [µ V2/Hz] was estimated for all frequencies between 1 and 30 Hz, then the relative power (%) was estimated for delta (1–4 Hz), theta (5–7 Hz), alpha (8–12 Hz), beta1 (13–15 Hz), and beta2 (15–30 Hz) frequency ranges. An example of EEG power spectra density is illustrated in Fig. 3. The relative power was represented using topographic maps: one map for each band in the AIRpre condition (1 min – 20 min) and seven maps for the O2 and AIRpost conditions, with seven intervals of analysis (1, 2, 5, 8, 10–12, 15–17, and 18–20 min). Minutes 18–20 from the EEG of subject no. 3, minute 5 from subject no. 8, and minute 2 from subject no. 9 were discarded due to loss of the EEG signal via Bluetooth.
2.4 Statistical analysis Analysis of variance (ANOVA) for repeated measures was applied to the relative power with the factor “condition” (AIRpre, O2, AIRpost); the sphericity assumption was assessed using Mauchly’s test. Greenhouse-Geisser epsilon adjustments for non-sphericity were applied where appropriate. A post-hoc paired sample two-tailed t-test adjusted for multiple comparisons with the Bonferroni method was used. Statistical significance was set at p < 0.05, corrected. Two-dimensional grand mean t-maps of the relative power were computed from the t-values to check the topographical distribution of the significance (Formaggio et al., 2008, 2013).
10
The t-maps were thresholded at p < 0.05 (|t| > 2.228) at 1, 8, 10–12, and 15–17 min (11 recordings) and thresholded at p < 0.05 (|t| > 1.812) at 2, 5, and 18–20 min (10 recordings). The different number of recordings is due to the loss of the EEG signal via Bluetooth in that time window.
3. Results
3.1 Group analysis While the subjects were breathing oxygen (2.8 ATA), the brain activity showed a fast and significant decrease in delta relative power, as compared to baseline, in the posterior regions starting from the early minutes of oxygen breathing (after 5 min), with a parallel and significant increase in the alpha rhythm in the same regions (Fig. 4). After decompression (AIRpost), as compared to baseline activity (AIRpre), the decrease in delta relative power was uniformly distributed over the scalp until minute 8. At 10–12 min, this decrease was principally localized in the posterior regions. The increase in alpha rhythm was uniformly distributed over the cerebral cortex until minute 8; this increase was still significant in the posterior regions until the end of the 20 minutes of recording (Fig. 4). The same pattern was also confirmed by the absolute power results (not reported here to avoid redundancy). ANOVA of the relative power of spectrum for each band disclosed a significant main effect for the factor condition (Table 1).
3.2 O2 vs. AIRpre During 2.8 ATA oxygen breathing, the delta relative power of the posterior electrodes started to decrease significantly from baseline (1 ATA air breathing) at 5 min, reaching the maximum decrease at 15–17 min that persisted until 18–20 min (Fig. 4). As shown on the t-maps (O2 vs. AIRpre), this significant decrease involved more regions over time (Fig. 5A). A similar pattern was observed in the theta band, with a
11
decrease in activity that persisted on the entire recording during oxygen breathing. The EEG relative power increase in the alpha band was evident in the same posterior regions, reaching an overall increase at 15–17 min. In the lower beta band, a similar pattern was observed in the posterior electrodes but without reaching statistical significance: the relative power markedly increased (principally over P3 and P4) after 5 min, reaching the maximum at 18–20 min. The relative power in the upper beta band increased significantly from baseline, particularly over the temporal electrodes (T3-T4-T5-T6).
3.3 AIRpost vs. AIRpre After decompression (air breathing), the delta relative power significantly decreased from baseline (1 ATA air breathing) over all electrodes until minute 8. At 10–12 min, the decrease was principally localized over the posterior regions. After 20 min, the power returned to the baseline value (Fig. 4). This change is shown on the t-maps (AIRpost vs. AIRpre) (Fig. 5B). This pattern was also observed in the theta band, with a decrease in activity that persisted until 5 min principally over the posterior regions, after which it returned to baseline. The alpha EEG relative power significantly increased in the same posterior regions during the first 2 minutes after decompression before returning to baseline at 20 min. The lower and upper beta bands were not involved during AIRpost.
4. Discussion The electrical activity of the brain is sensitive to its oxygen supply. On high-altitude exposure, for example, EEG topographical changes are remarkable for an increase in delta and theta activities and a decrease in alpha activity (Papadelis et al., 2007). Furthermore, because EEG is profoundly affected by transient hypoxia (Ozaki et al., 1995) it can be considered as an indicator of cerebral metabolism (Jordan 1993). Hyperventilation is also known to cause, especially in children, an increase in synchronous EEG slow activity, together with a slowing of fast rhythms (Niedermeyer and Lopes da Silva 2000) and vasoconstriction of the cerebral resistance vessels (Visser et al., 1996b). However, the effects of oxygen
12
exposure on cerebral activity are not well known. Few studies have quantified by EEG the effects of O2 toxicity on cerebral activity (Visser et al., 1996a,b). They showed a global increase in power, which occurred in the first part of the experiment and reached a steady state after 100% O2 inhalation was started. An increase in high-frequency alpha activity and a non-correlated decrease in low frequency alpha were found after decompression. The gradual increase in low-frequency alpha activity toward the end of the experiment was not necessarily an effect of oxygen but instead might have been a slow effect of compression. Nonetheless, the results of this study do not allow to draw definitive conclusions, particularly in the subject who reported a seizure. Our findings, associated to a time-dependent decrease in fast delta activity in the posterior regions and an increase in alpha activity in the same regions, showed a behaviour opposite that of the hypoxia condition. Since delta activity is a sign that the cortical inhibitory mechanisms (e.g., the circuit of pyramidal and stellate cells) are more active, the delta decrease observed during oxygen breathing could represent reduced performance of cortical inhibitory mechanisms. During prolonged exposure to HBO in these professional divers, EEG changes mainly in the delta and alpha ranges, albeit with small latencies, were reproducible among the subjects. The hypothesis for vasomotor cerebral activity is very attractive. This phenomenon would consist of vasoconstriction followed by vasodilatation. The opposing behaviour (increase in delta activity and decrease in the alpha band) described by Manganotti et al. after transcranial magnetic stimulation (TMS) over the cerebral motor area (Manganotti et al., 2012) and the reduction in cerebral vasomotor reactivity after TMS stimulation observed by Vernieri et al. (2009) is consistent with this hypothesis. A pattern of vasoconstriction and vasodilatation during oxygen breathing previously reported in animals (Torbati et al., 1979) may be related to complex nitric oxide metabolism (Demchenko et al., 2000) and the role of oxygen free radicals (Elayan et al., 2000). Nitric oxide (NO) is a key intercellular
13
messenger in cerebral and peripheral hemodynamics. It plays a central role in regulating cerebral blood flow and cell viability and in protecting nerve cells or fibers against pathogenic factors associated with cerebral ischemia, trauma, and hemorrhage. Specifically, cerebral blood flow is increased and cerebral vascular resistance is decreased by NO derived from endothelial cells, autonomic nitrergic nerves, or brain neurons under resting and stimulated conditions (Toda et al., 2009). We would assume that during oxygen breathing vasoconstriction is followed by vasodilatation with an increase in alpha activity in the posterior regions, independently of oxygen pressure (Weaver and Howe, 1992). Vascular response to hyperoxia leads to a reduction in cerebral blood flow. In the present study, the hyperoxic conditions produced a residual effect on the CNS, with a significant alteration of both slow and fast rhythms until 8 min of the AIRpost stage. Neuron functional connectivity due to hyperoxia may be responsible for the observed increase in the relative alpha power in the posterior regions. Finally, we observed that the significant increase in alpha activity persisted until 20 minutes of EEG recording only in the posterior regions. Despite strict wakefulness maintenance while recording, we cannot completely exclude that a slight increase in theta components as a marker of increased drowsiness may have occurred. Nonetheless, our results showed a decrease in theta activity, especially during O2 breathing. While many mobile EEG systems have been developed in different contexts and for a variety of purposes (i.e., in epilepsy, sleep, brain computer interface) (Askamp and van Putten, 2013), here we focus on their potential application in hyperbaric chambers. Electroencephalographic studies in hyperbaric chamber are not common. The main limitations are the problems with recording the EEG signal owing to the fact that the chamber has a steel casing, i.e., a Faraday cage, and that the power supply of the equipment inside the chamber cannot be powered by alternating current because it would generate sparks. The development of Holter apparatuses, i.e., devices that amplify signals close to the subject, allows for using low-voltage direct current (DC) power supply, which does not generate dangerous sparks during circuit opening and closing. Furthermore, amplification near the signal source allows for the use of short
14
cables thus minimizing the eddy currents. With the advent of Bluetooth technology, cables have been replaced with low-length radio waves that allow its use in all types of hyperbaric chambers. The main advantage of Bluetooth technology is the improved EEG signal quality over older technologies, thus reducing environmental or cable movement interferences. Oxygen toxicity is a problem in diving, with potentially fatal consequences for the diver in the water. When divers use a closed-circuit oxygen rebreathing apparatus, they are breathing O2 100% and are exposed to hyperoxide conditions. In this scenario, the body could be affected in different ways depending on the type of exposure. Central nervous system toxicity is caused by short exposure to high partial pressures of oxygen at greater than atmospheric pressure. Pulmonary and ocular toxicity result from longer exposure to elevated oxygen levels at normal pressure. Exposure to HBO can cause oxidative damage to cell membranes, collapse of the pulmonary alveoli, retinal detachment, and seizures (Acott 1999; Beehler, 1964, Alcaraz-García et al., 2008). Symptoms may include disorientation, breathing problems, and vision changes such as myopia. Since symptoms of visual disturbance, ear problems, dizziness, confusion, and nausea can all arise from such common factors as narcosis, congestion and coldness, the diagnosis of CNS O2T in divers prior to seizure is difficult to establish. The findings of this study, i.e., significant changes in the relative power of both delta and alpha oscillations between oxygen exposure and the baseline stage, suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygen-toxicity diving-related problems”. The approach to monitoring cerebral power spectra could be useful to evaluate the physiological effects of oxygen breathing and to delineate different diving profiles, providing indications on the potential risk of a dive. With this study we have constructed a normative dataset of EEG responses to HBO, to which the EEGs of “divers experiencing problems" can be compared statistically. Training activity in hyperbaric chambers during EEG measurements can be very important in the selection of divers who are more suitable to deal with a stressful undersea environment. We are also planning in future studies to include EEG evaluation during oxygen breathing at a pressure of 1 ATA and air at 2.8 ATA in order to determine how oxygen and hyperbaric pressure variables affect
15
bioelectrical activity in human professional divers.
Conflict of interest The authors report no conflicts of interest.
Acknowledgements This study was funded by grant SMD L-023: Rilevazioni elettrofisiologiche in immersione; Cap. 1322, Italian Ministry of Defense, Direzione Generale della Sanità Militare, 2010.
References
Acott C. Oxygen Toxicity – A brief history of Oxygen in Diving. SPUMS J 1999; 29:150-155.
16
Alcaraz-García MJ, Albaladejo MD, Acevedo C, Olea A, Zamora S, Martínez P, et al. Effects of hyperoxia on biomarkers of oxidative stress in closed-circuit oxygen military divers. J Physiol Biochem 2008;6 4:135-41.
Arieli R, Shochat T, Adir Y. CNS toxicity in closed-circuit oxygen diving: symptoms reported from 2527 dives. Aviat Space Environ Med 2006; 77:526-32.
Babiloni C, Vecchio F, Lizio R, Ferri R, Rodriguez G, Marzano N, et al. Resting state cortical rhythms in mild cognitive impairment and Alzheimer's disease: electroencephalographic evidence. J Alzheimers Dis 2011; 26:201-14.
Beehler, CC. Oxygen and the eye. Survey of Ophthalmology 1964; 45: 549–60.
Bitterman N, Melamed Y, Perlman I. CNS oxygen toxicity in the rat: role of ambient illumination. Undersea Biomed Res 1986; 13:19-25.
Butler FK Jr, Knafelc ME. Screening for oxygen intolerance in U.S. Navy divers. Undersea Biomed res 1986; 13:91-98.
Butler FK Jr, Thalmann ED. Central nervous system oxygen toxicity in closed circuit scuba divers II. Undersea Biomed Res 1986; 13:193-223.
17
Cohn R, Gersch I. Changes in brain potentials during convulsions induced by oxygen under pressure. J Neurophysiol 1945; 8:155-60.
Demchenko IT, Boso AE, O’Neill TJ, Bennett PB, Piantadosi CA. Nitric oxide and cerebral blood flow responses to hyperbaric oxygen. J Appl Physiol 2000; 88:1381-9.
Donald KW. Oxygen poisoning in man. Br Med J 1947; 1:667-72.
Elayan IM, Axley MJ, Prasad PV, Ahlers ST, Auker CR. Effect of hyperbaric oxygen treatment on nitric oxide and oxygen free radicals in rat brain. J Neurophysiol 2000; 83:2022-9.
Formaggio E, Storti SF, Avesani M, Cerini R, Milanese F, Gasparini A, et al. EEG and fMRI coregistration to investigate the cortical oscillatory activities during finger movement. Brain Topogr 2008; 21:100-11.
Formaggio E, Storti SF, Boscolo Galazzo I, Gandolfi M, Geroin G, Smania N, et al. Modulation of eventrelated desynchronization in robot-assisted hand performance: brain oscillatory changes in active, passive and imagined movements. J Neuroeng Rehabi 2013; doi: 10.1186/1743-0003-10-24.
18
Harabin AL. Human central nervous system oxygen toxicity data from 1945 to 1986. Bethesda, MD: Naval Medical Research Institute 1993; (NMRI Report 93-03).
Ibner C, Ancoli-Israel S, Chesson A, Quan SF (eds). The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specifications. 1st edition. Westchester, III: American Academy of Sleep Medicine; 2007.
Imlay JA. Pathways of oxidative damage. Annu Rev Microbiol 2003; 57:395-418.
Jordan KG. Continuous EEG and evoked potential monitoring in the neuroscience intensive care unit. J Clin Neurophysiol 1993; 10:445-75.
Lemaire C, Rostain JC. The High Pressure Nervous Syndrome and Performance. Marseille: Octares, 1988.
Manganotti P, Formaggio E, Storti SF, De Massari D, Zamboni A, Bertoldo A, et al. Time-frequency analysis of short-lasting modulation of EEG induced by intracortical and transcallosal paired TMS over motor areas. J Neurophysiol 2012; 107:2475-84.
Morris LE, Knott JR, Pittinger CB. Electroencephalographic and blood gas analysis in human surgical patients during xenon anesthesia. Anesthesiology 1955; 16:312-9.
19
Murata M, Suzuki M, Hasegawa Y, Nohara S, Kurachi M. Improvement of occipital alpha activity by repetitive hyperbaric oxygen therapy in patients with carbon monoxide poisoning: a possible indicator for treatment efficacy. J Neurol Sci 2005; 235:69-74.
Niedermeyer E, Lopes da Silva FH. Electroencephalography: Basic Principles, Clinical Applications, and Related Fields. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 2000.
Neubauer RA. High dose oxygen therapy in stroke. Eur J Neurol 2000; 7:151.
Okuda S, Matsuoka S, Mohri M. Topographic electroencephalographic studies in a hyperbaric environment--specific reference to high pressure nervous syndrome. J UOEH. 1988;10:247-61.
Ozaki H, Watanabe S, Suzuki H. Topographic EEG changes due to hypobaric hypoxia at simulated high altitude. Electroencephalogr Clin Neurophysiol 1995; 94:349-56.
Papadelis C, Kourtidou-Papadeli C, Bamidis PD, Maglaveras N, Pappas K. The effect of hypobaric hypoxia on multichannel EEG signal complexity. Clin Neurophysiol 2007; 118:31-52.
20
Pastena L, Mainardi G, Faralli F, Gagliardi R. Analysis of cerebral bioelectrical activity during the compression phase of a saturation dive. Aviat Space Environ Med 1999; 70:270-6.
Pastena L, Faralli F, Mainardi G, Gagliardi R. EEG patterns associated with nitrogen narcosis (breathing air at 9 ATA). Aviat Space Environ Med 2005; 76:1031-6.
Piantadosi CA, Clinton RL, Thalmann ED. Prolonged oxygen exposures in immersed exercising divers at 25 fsw (1.76 ATA). Undersea Biomed Res 1979; 6:347-56.
Rostain JC, Charpy JP. Effects upon the EEG of psychometric performance during deep dives in heliumoxygen atmosphere. Electroencephalogr Clin Neurophysiol 1976; 40:571-84.
Storti SF, Formaggio E, Franchini E, Bongiovanni LG, Cerini R, Fiaschi A, et al. A multimodal imaging approach to the evaluation of post-traumatic epilepsy. Magn Reson Mater Phy 2012; 25:345-60.
Takahashi TOY, Okamoto K, Yoshimura Y, Matsuoka Y, Oomori N. The current status of carbon monoxide poisoning and treatment with hyperbaric oxygen. Rinsho Seishin Igaku 1998;27:701–6.
Toda N, Ayajiki K, Okamura T. Cerebral blood flow regulation by nitric oxide: recent advances. Pharmacol Rev 2009; 61:62-97.
21
Torbati D, Parolla D, Lavy S. Organ blood flow, cardiac output, arterial blood pressure and vascular resistance in rats exposed to various oxygen pressures. Aviat Space Environ Med 1979; 50:256-63.
Vernieri F, Maggio P, Tibuzzi F, Filippi MM, Pasqualetti P, Melgari JM, et al. High frequency repetitive transcranial magnetic stimulation decreases cerebral vasomotor reactivity. Clin Neurophysiol 2009; 120:1188-94.
Visser GH, van Hulst RA, Wieneke GH, van Huffelen AC. The contribution of conventional and quantitative electroencephalography during monitoring of exposure to hyperbaric oxygen. Undersea Hyperb Med 1996a; 23:91-8.
Visser GH, Van Hulst RA, Wieneke GH, Van Huffelen AC. Transcranial Doppler sonographic measurements of middle cerebral artery flow velocity during hyperbaric oxygen exposures. Undersea Hyperb Med 1996; 23:157-65.
Weaver LK, Hopkins RO, Chan KJ, Churchill S, Elliott CG, Clemmer TP, et al. Hyperbaric oxygen for acute carbon monoxide poisoning. N Engl J Med 2002; 347:1057-67.
22
Weaver LK, Howe S. Normobaric measurement of arterial oxygen tension in subjects exposed to hyperbaric oxygen. Chest 1992; 102:1175-81.
Zal’tsman GL. Physiological principles of sojourn of a human in conditions of raised pressure of gases medium [in Russian]. Ohio: Foreign Technology Division, Wright Patterson Air Force Base 1967; AD 655, 360.
23
Figure legends
Figure 1. (A) 32-channel HydroCel Geodesic Sensor Net Map. (B) A navy diver is wearing the EEG cap outside the hyperbaric chamber. (C) During the experiment, the subject is reclined on a cot with eyes closed. The 32-channel EEG system is connected via Bluetooth wireless transmission to a notebook that visualizes the EEG signal outside the hyperbaric chamber.
Figure 2. (A) Dive profile and recording sessions. The subject is reclined on a cot with eyes closed. A 20minute baseline EEG recording was made at 1 ATA breathing air (air) in the open hyperbaric chamber. In the closed chamber, a 2-min compression profile (descent rate 9 m min-1) breathing air was used to reach the next stage at a pressure of 2.8 ATA (compression). At this pressure, the subject breathed pure oxygen via an oronasal mask (oxygen) and a 20-min EEG was acquired. After decompression, back on air breathing, the EEG of each subject was recorded for 20 min, discarding the first 2 min (ascent rate 9 m min-1) (decompression). (B) Example of an EEG signal after preprocessing: air (AIRpre), end of oxygen breathing at minute 20 (O2), and end of air breathing after decompression at minute 20 (AIRpost).
Figure 3. Power spectral densities (μV2/Hz) of the EEG signal (only 19 channels visualized in the international 10-20 system positions) obtained by FFT on 2 s epochs, in one professional diver. (A) O2 min 18-20 (green) vs. AIRpre average of 20-minute EEG recording (blue). (B) AIRpost min 8 (red) vs. AIRpre average of 20-minute EEG recording (blue). The y-axis has a different range for each channel in order to capture the differences between the spectra. The scale is the same in A and B.
24
Figure 4. Relative EEG power maps in delta, theta, alpha, lower beta and upper beta bands averaged from 11 subjects. During air breathing averaged in 20 min (AIRpre), during O2 breathing (1 ATA) (O2) and during air breathing after decompression (AIRpost) at 1, 2, 5, 8, 10–12, 15–17, 18–20 min. Red coding indicates maximal relative power. At 1, 2, 5, 8 min after decompression there is a transient phase where pressure and oxygen effects take some time to stabilize (e.g., saturation effect).
Figure 5. T-maps O2 vs. AIRpre (A) and AIRpost vs. AIRpre (B) of relative EEG power in delta, theta, alpha, beta1 and beta2 bands. Each minute of EEG recording during O2 and AIRpost was compared to the entire AIRpre recording (average of 20 minutes), respectively. The threshold is p < 0.05 (|t| > 2.228) at 1, 8, 10–12 and 15–17 min (11 recordings) and p < 0.05 (|t| > 1.812) at 5 and 18–20 min (10 recordings). (+) indicates significance for Bonferroni correction. The effects in 'oxygen vs. air' and ‘after decompression vs. air’ show transients in opposite directions.
25
Table 1. Analysis of variance (ANOVA) for repeated measures applied to the relative power with the factor “condition” (AIRpre, O2, AIRpost).
ANOVA factor “condition” (AIRpre, O2, AIRpost) 1'
2'
5'
8'
10'-12'
15'-17'
18'-20'
F(2,20) = 20.397
F(2,18) = 18.422
F(2,18) = 14.933
F(2,20) = 15.605
F(2,20) = 7.131
F(2,20) = 6.064
F(2,18) = 4.897
p < 0.001
p < 0.001
p < 0.001
p < 0.001
p < 0.01
p < 0.01
p < 0.05
F(2,20) = 7.650
F(2,18) = 7.881
F(2,18) = 4.866
F(2,20) = 8.436
F(2,20) = 7.856
F(2,18) = 6.769
p < 0.005
p < 0.005
p < 0.01
F(2,20) = 3.931
F(2,20) = 5.259
δ
θ
n.s. p < 0.005
p < 0.005
p < 0.05
F(2,20) = 15.592
F(1.304, 11.732) = 16.725
F(2,18) = 9.217
p < 0.001
p < 0.005
p < 0.005
p < 0.01
p < 0.05
p < 0.05
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
F(2,18) = 4.629
F(2,20) = 3.737
F(2,20) = 4.569
α
β1
F(2,20) = 4.432
β2
F(2,20) = 7.043
n.s.
n.s. p < 0.05
n.s.
F(2,18) = 4.207 n.s.
p < 0.05
p < 0.05
p < 0.05
p < 0.05
n.s. = not significant
26
27
28
30
31
Abstract Objective: The aim was to investigate and define possible alterations in cerebral activity during prolonged hyperbaric oxygen exposure and decompression as compared to baseline activity. Methods: Thirty-two channel electroencephalography (EEG) was recorded with a Bluetooth EEG system in 11 subjects. A 20-minute EEG recording was carried out under three different conditions: breathing air inside a hyperbaric chamber at sea level; breathing oxygen at a simulated depth of 18 msw; breathing air at sea level after decompression. Relative EEG power was estimated in frequency ranges. Results: During oxygen breathing, brain activity showed an early fast delta decrease in the posterior regions, with a synchronous and significant increase in alpha in the same regions. After decompression, the delta relative power decrease was uniformly distributed over the cerebral cortex until minute 8, and the alpha relative power was maximal in the posterior regions during the first 2 minutes. Conclusions: These results may be relevant for establishing a reference point in future studies on oxygen-sensitive subjects who reported problems during oxygen diving. Significance: Significant changes in EEG relative power suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygen-toxicity diving-related problems".
32
S1388-2457(14)00288-0 http://dx.doi.org/10.1016/j.clinph.2014.05.013 CLINPH 2007110
To appear in:
Clinical Neurophysiology
Accepted Date:
5 May 2014
Please cite this article as: Pastena, L., Formaggio, E., Storti, S.F., Faralli, F., Melucci, M., Gagliardi, R., Ricciardi, L., Ruffino, G., Tracking EEG changes during the exposure to hyperbaric oxygen, Clinical Neurophysiology (2014), doi: http://dx.doi.org/10.1016/j.clinph.2014.05.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Tracking EEG changes during the exposure to hyperbaric oxygen
Lucio Pastenaa, Emanuela Formaggiob,§, Silvia Francesca Stortic,§,*, Fabio Farallid, Massimo Meluccid, Riccardo Gagliardid, Lucio Ricciardid, Giovanni Ruffinod
a
Department of Neurological Sciences, University of Rome, La Sapienza, Rome, Italy
b
Department of Neurophysiology, Foundation IRCCS San Camillo Hospital, Venice, Italy
c
Department of Neurological and Movement Sciences, University of Verona, Verona, Italy
d
Italian Navy Medical Service Comsubin Varignano, Le Grazie (La Spezia), Italy
§
These authors equally contributed to the study.
1
* Corresponding author: Silvia Francesca Storti Dipartimento di Scienze Neurologiche e del Movimento Policlinico G.B. Rossi – P.le L.A. Scuro 10 – 37134 Verona, Italy Tel: +39 045-8126394 Fax: +39045-8027276 E-mail: [email protected] Abstract Objective: The aim was to investigate and define possible alterations in cerebral activity during prolonged hyperbaric oxygen exposure and decompression as compared to baseline activity. Methods: Thirty-two channel electroencephalography (EEG) was recorded with a Bluetooth EEG system in 11 subjects. A 20-minute EEG recording was carried out under three different conditions: breathing air inside a hyperbaric chamber at sea level; breathing oxygen at a simulated depth of 18 msw; breathing air at sea level after decompression. Relative EEG power was estimated in frequency ranges. Results: During oxygen breathing, brain activity showed an early fast delta decrease in the posterior regions, with a synchronous and significant increase in alpha in the same regions. After decompression, the delta
2
relative power decrease was uniformly distributed over the cerebral cortex until minute 8, and the alpha relative power was maximal in the posterior regions during the first 2 minutes. Conclusions: These results may be relevant for establishing a reference point in future studies on oxygen-sensitive subjects who reported problems during oxygen diving. Significance: Significant changes in EEG relative power suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygen-toxicity diving-related problems".
Keywords: EEG brain mapping; oxygen toxicity; brain activity; EEG frequency analysis.
Highlights -
We investigated EEG changes in professional divers after oxygen breathing at a simulated depth of 18 msw and after decompression breathing air at sea level.
-
A significant decrease in delta and an increase in alpha relative power in the posterior regions were observed during oxygen breathing.
-
The results suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygentoxicity diving-related problems".
1. Introduction The use of oxygen mixture and closed circuit apparatus in diving at various depths augments the risk of central nervous system oxygen toxicity (CNS O2T). Oxygen poisoning can manifest itself with symptoms ranging from nausea, dizziness, hearing and visual disturbance, amnesia, vertigo, irritability, localized
3
muscular twitching, to loss of consciousness, tonic clonic seizures, and EEG alterations such as an increase in theta waves, a decrease in alpha waves, microsleep and a decline in psychometric performance. Prolonged exposure to hyperbaric oxygen (HBO) will result in permanent paralysis and death (Donald 1947). To prevent this syndrome, prompt recognition of the earliest symptoms, as described after dry or humidified HBO exposure, is vital (Donald 1947; Piantadosi et al., 1979; Butler and Thalmann 1986). An overview of work in this area is given by Harabin and Donald (Donald 1947; Harabin 1993). More recently, Arieli et al. reviewed 2527 dive reports to analyze the relationship between symptom onset and its dependence on depth and dive time (Arieli et al., 2006). In some animals, environmental factors, such as cold water and darkness, can increase oxygen sensitivity (Bitterman et al., 1986). During the selection of military divers, an oxygen tolerance test (OTT) can be used to identify subjects with susceptibility to CNS O2T (Butler and Knafelc 1986; Butler and Thalmann 1986). The Italian Navy OTT procedure involves exposing divers to 2.8 atmosphere absolute pressure (ATA) for 15 min in a dry chamber breathing 100% O2 through a face mask. The electroencephalogram (EEG) pattern is known to significantly change during saturation dives (Lemaire and Rostain 1988; Okuda et al., 1988 Pastena et al., 1999), but few researchers have studied EEG modifications in humans during oxygen breathing in a hyperbaric chamber. The main contribution to the study of CNS O2T was made by Visser et al. (1996a,b). Quantitative EEG findings are described for a group of 23 subjects during HBO exposure and for one subject who had a generalized tonic-clonic seizure on exposure to HBO. The subjects were exposed to 100% O2 for 30 min at 2.8 bar. EEG changes were minor and not considered indicative of a HBO effect on the brain in the group of subjects who showed no signs of toxicity. Pre-convulsive EEG changes were detected in the subject who experienced a seizure but they were too insignificant and did not clearly herald clinical signs (Visser et al., 1996a). In the same group of subjects, the blood flow velocity in the right middle cerebral artery was monitored by transcranial Doppler (TCD). The TCD mean velocity
4
decreased during HBO exposure probably due to vasoconstriction of the cerebral resistance vessels, whereas it increased in the subject who experienced a generalized seizure. Importantly, however, no definitive conclusions can be drawn owing to the small number of subjects with toxicity in this study (Visser et al., 1996b). Stronger EEG abnormalities can be found already in the compression phase. For example, during the compression phase of a saturation dive reaching 250 msw and breathing O2 at different pressures (pO2 0.44–0.48 ATA), an increase in theta and beta rhythms in the central regions of the brain were observed during the entire compression period, whereas an increase in delta was noted at a depth of 100 msw (Pastena et al., 1999). In extremely stressful situations, a theta increase could be considered an expression of altered activity in the frontal central areas while maintaining cerebral homeostasis in the presence of remarkable environmental stressors. The increase in delta rhythm could be the epiphenomenon of partial cortical deafferentation in the posterior areas. Rostain and Charpy studied EEG modifications and psychometric performance in healthy subjects during two simulated dives at 500 and 610 m with a helium-oxygen-breathing mixture (pO2 0.4 ATA). The EEG modifications considered as high-pressure nervous syndrome (HPNS) began at 300 m. They were remarkable for an increase in slow activity (particularly theta) in the anterior and middle regions of the scalp, depression of fast activity, and transformation of the waking EEG into one resembling that of stage I sleep. In addition, the sensorimotor performance scores on psychometric testing were lower in all subjects (Rostain and Charpy 1976). Changes in electrical cortical activity have been observed during convulsions induced by an increase in oxygen pressure (Cohn and Gersch 1945). Research into the prevention of CNS O2T has been informed by EEG findings in animals during the preconvulsive period of HBO exposure; however, these studies described only the onset of “irritative” figures (spikes, polyspikes) in many subjects, without any
5
definitive results. Today, EEG brain mapping combined with fast Fourier transform analysis can reveal variations in bioelectrical activities associated with the environmental factor under study, for example, changes in EEG cortical activity during hyperbaric exposure (Zal’tsman 1967), xenon anaesthesia (Morris et al., 1955), saturation dives (Pastena et al., 1999), and nitrogen narcosis (Pastena et al., 2005). Hyperbaric oxygen treatment is indicated for a wide range of illnesses (Neubauer 2000; Murata et al., 2005). Murata et al. evaluated the efficacy of repetitive HBO treatments by monitoring quantitative EEG for symptoms of acute carbon monoxide poisoning and prevention of delayed neuropsychiatric sequelae. Examining the effects of repetitive HBO treatment on quantitative EEG in the patients with CO poisoning, they found that the peak alpha frequency and the relative alpha power significantly increased in the occipital regions after repetitive HBO treatments (Murata et al., 2005). From these findings, the authors hypothesized that monitoring of peak alpha frequency may be an objective indicator for the efficacy of HBO treatments to prevent delayed neuropsychiatric sequelae. Although repetitive HBO therapy is known to be effective for their prevention, the precise mechanisms of its action for CO intoxication are not completely understood (Weaver et al., 2002, Takahashi et al., 1998). Furthermore, hyperbaric oxygenation once played an integral role in the treatment of early stroke, where changes in the EEG pattern observed during treatment (Neubauer 2000) aided in guiding neurorehabilitation. Though widely applied in the clinical setting, the effects of oxygen therapy on healthy subjects are not fully understood (Imlay 2003). Quantitative EEG plays a significant role in EEG-based clinical diagnosis and studies of brain function, revealing new possibilities for clinical application in cognitive and behavioural areas (Babiloni et al., 2011; Storti et al., 2013). The quantification of changes in brain rhythms during oxygen breathing is essential to determine a basic pattern in healthy and trained divers in order to detect alterations in subjects at risk of CNS O2T. EEG quantification in normative data could
6
be used as a reference point for future studies that would serve as a landmark of oxygen-induced brain activity and could be useful in the medical treatment of divers reporting problems. Therefore, the aim of this study was to determine how oxygen, assumed at a constant hyperbaric pressure of 2.8 ATA, affects the bioelectrical activity in human professional divers during the compression phase, and to define the EEG patterns of professional divers during this condition as compared to pre- and post-oxygen breathing. One-hour EEGs were recorded in 11 subjects in order to quantify brain activity modifications under three different conditions: air (baseline at sea level), oxygen breathing (at a simulated depth of 18 msw), and air (after decompression).
2. Methods
2.1 EEG with Bluetooth technology The EEG was recorded inside a hyperbaric chamber using a Holter apparatus equipped with Bluetooth wireless technology. Hyperbaric chambers are equipped with portholes, made of transparent resins (i.e., pressure-resistant Plexiglas) through which Bluetooth radio waves are transmitted outside the chamber. The EEG system, located outside the chamber, should be placed next to the porthole in order to obtain more stable acquisition. The EEG system (Ates Medica Device, Verona, Italy) is a portable recorder connected via Bluetooth wireless transmission to a notebook that visualizes the EEG signal (Fig. 1). EEG recordings were performed using 32 Ag/AgCl electrodes (Electrical Geodesic, Inc., Eugene, OR), which are soaked in a physiological-potassium solution for about 5 min before the beginning of the recording. This solution allows a microclimate to develop between the subject’s scalp and the electrode and keeps the impedances low even during long EEG recordings. The net was adjusted so that the 18, Cz, Oz and pre-auricular points were correctly placed according to the international 10/20 system. By virtue of the net’s geodesic
7
tension structure, all electrodes were evenly distributed over the scalp at approximately the same location in all 11 subjects. The data were recorded against a vertex electrode reference (Cz) at a sampling rate of 250 Hz using the software package Geodesic EEG System on Neurotravel technology (Ates Medica Device and Electrical Geodesic, Inc.). An anti-aliasing hardware low-pass filter with a cut-off frequency of 96 Hz was applied. Electrode impedances were verified and kept below 30 kΩ.
2.2 Subjects and experimental protocol The study sample was 11 healthy male navy divers (mean age 46.2 ±4.9 years). Exclusion criteria were a medical history of respiratory problems, sleep disturbances, smokers, and overweight. All subjects were experienced divers with at least 15 years of experience. All had undergone OTTs and none had ever shown signs of CNS toxicity, neither during an OTT nor during operational diving with 100% O2, nitrox or other mixtures. Four belong to the Operational Divers Group and are personnel specialized in free diving with air to 60 msw, nitrox to 54 msw, heliox to 150 msw, and to 250 msw with a minisubmarine or a special suit. The divers in the moderately experienced group possess a license for diving to 60 msw. The internal ethics committee of University approved the experimental protocol and oral and written informed consent was obtained from the subjects before participating in the study. A simulated depth of 18 msw (2.8 ATA) was selected for the study because exposure at this breathing depth is a routine part of navy diver training courses. The hyperbaric chamber used in the study complied with Italian Navy standards for safety equipment and emergency procedures. All subjects were studied individually. They were accompanied during the dive by a technician who assisted the subject if needed and were monitored by closed circuit television. Arousal was maintained throughout the whole recording session by administering an external acoustic stimulation as soon as one 30-second epoch showed a reduction of more than 50% of the background alpha rhythm, according
8
to American Academy of Sleep Medicine scoring rules (Ibner et al., 2007). The procedure was carried out by an experienced neurophysiologist who was also responsible for detecting any epileptiform phenomena. Visual inspection of EEG signals showed no abnormalities. The divers had complete saturation of blood haemoglobin; nonetheless, the true value of pO2, which requires arterial sampling, cannot be performed inside a hyperbaric chamber. Each recording session lasted 20 min, during which the subject reclined on a cot with eyes closed. A baseline 20-min EEG recording was made at 1 ATA breathing air (AIRpre) in an open chamber. A 2-min compression profile (descent rate 9 m min-1) breathing air was used to reach the oxygen stage at a pressure of 2.8 ATA. At this pressure, the subject breathed pure oxygen via the oronasal mask (O2) and a second 20-min EEG recording was acquired. The atmosphere within the hyperbaric chamber was controlled to maintain the total pressure of 2.8 ATA. After the decompression stage, back on air breathing, the EEG of each subject was recorded for 20 min (AIRpost), discarding the first 2 min (ascent rate 9 m min-1). The overall time of EEG recording was about 1 hour (Fig. 2) The chamber was compressed with air, and a separate breathing circuit was used for the oxygen stage. As the chamber was equipped with to a continuous ventilation system, eventual oxygen leaks from the mask were not thought to influence the atmosphere inside the hyperbaric chamber.
2.3 Data analysis The data were processed in Matlab 7 (MathWorks, Natick, MA) using scripts based on EEGLAB 6.01 (EEGLAB toolbox; http://www.sccn.ucsd.edu/eeglab), as well as a dedicated home-made code created especially for this study. The EEG recordings were band-pass filtered from 1 to 30 Hz with a finite impulse response filter. Visible artifacts in the EEG recordings (i.e., eye movements, cardiac activity, and scalp muscle contraction) were removed using an independent component analysis
9
procedure. Electrode 18 (positioned on the nasion) was rejected because of ocular artifact. The data were processed using an average reference. The EEG data were divided into different time intervals for each condition: one interval for the AIRpre condition (1– 20 min) and seven intervals for the O2 and AIRpost conditions (1, 2, 5, 8, 10–12, 15–17, and 18–20 min). Since these time intervals are sufficient to describe the complete phenomenon, the analysis and the representation of the power maps were limited to these windows. The EEG data for each interval were divided into epochs of 2 s. A fast Fourier transform was applied to non-overlapping epochs, each containing 500 data points for all electrodes and the three experimental conditions, and then averaged across epochs under the same interval condition. The recordings were Hamming-windowed to control for spectral leakage. The power spectra density [µ V2/Hz] was estimated for all frequencies between 1 and 30 Hz, then the relative power (%) was estimated for delta (1–4 Hz), theta (5–7 Hz), alpha (8–12 Hz), beta1 (13–15 Hz), and beta2 (15–30 Hz) frequency ranges. An example of EEG power spectra density is illustrated in Fig. 3. The relative power was represented using topographic maps: one map for each band in the AIRpre condition (1 min – 20 min) and seven maps for the O2 and AIRpost conditions, with seven intervals of analysis (1, 2, 5, 8, 10–12, 15–17, and 18–20 min). Minutes 18–20 from the EEG of subject no. 3, minute 5 from subject no. 8, and minute 2 from subject no. 9 were discarded due to loss of the EEG signal via Bluetooth.
2.4 Statistical analysis Analysis of variance (ANOVA) for repeated measures was applied to the relative power with the factor “condition” (AIRpre, O2, AIRpost); the sphericity assumption was assessed using Mauchly’s test. Greenhouse-Geisser epsilon adjustments for non-sphericity were applied where appropriate. A post-hoc paired sample two-tailed t-test adjusted for multiple comparisons with the Bonferroni method was used. Statistical significance was set at p < 0.05, corrected. Two-dimensional grand mean t-maps of the relative power were computed from the t-values to check the topographical distribution of the significance (Formaggio et al., 2008, 2013).
10
The t-maps were thresholded at p < 0.05 (|t| > 2.228) at 1, 8, 10–12, and 15–17 min (11 recordings) and thresholded at p < 0.05 (|t| > 1.812) at 2, 5, and 18–20 min (10 recordings). The different number of recordings is due to the loss of the EEG signal via Bluetooth in that time window.
3. Results
3.1 Group analysis While the subjects were breathing oxygen (2.8 ATA), the brain activity showed a fast and significant decrease in delta relative power, as compared to baseline, in the posterior regions starting from the early minutes of oxygen breathing (after 5 min), with a parallel and significant increase in the alpha rhythm in the same regions (Fig. 4). After decompression (AIRpost), as compared to baseline activity (AIRpre), the decrease in delta relative power was uniformly distributed over the scalp until minute 8. At 10–12 min, this decrease was principally localized in the posterior regions. The increase in alpha rhythm was uniformly distributed over the cerebral cortex until minute 8; this increase was still significant in the posterior regions until the end of the 20 minutes of recording (Fig. 4). The same pattern was also confirmed by the absolute power results (not reported here to avoid redundancy). ANOVA of the relative power of spectrum for each band disclosed a significant main effect for the factor condition (Table 1).
3.2 O2 vs. AIRpre During 2.8 ATA oxygen breathing, the delta relative power of the posterior electrodes started to decrease significantly from baseline (1 ATA air breathing) at 5 min, reaching the maximum decrease at 15–17 min that persisted until 18–20 min (Fig. 4). As shown on the t-maps (O2 vs. AIRpre), this significant decrease involved more regions over time (Fig. 5A). A similar pattern was observed in the theta band, with a
11
decrease in activity that persisted on the entire recording during oxygen breathing. The EEG relative power increase in the alpha band was evident in the same posterior regions, reaching an overall increase at 15–17 min. In the lower beta band, a similar pattern was observed in the posterior electrodes but without reaching statistical significance: the relative power markedly increased (principally over P3 and P4) after 5 min, reaching the maximum at 18–20 min. The relative power in the upper beta band increased significantly from baseline, particularly over the temporal electrodes (T3-T4-T5-T6).
3.3 AIRpost vs. AIRpre After decompression (air breathing), the delta relative power significantly decreased from baseline (1 ATA air breathing) over all electrodes until minute 8. At 10–12 min, the decrease was principally localized over the posterior regions. After 20 min, the power returned to the baseline value (Fig. 4). This change is shown on the t-maps (AIRpost vs. AIRpre) (Fig. 5B). This pattern was also observed in the theta band, with a decrease in activity that persisted until 5 min principally over the posterior regions, after which it returned to baseline. The alpha EEG relative power significantly increased in the same posterior regions during the first 2 minutes after decompression before returning to baseline at 20 min. The lower and upper beta bands were not involved during AIRpost.
4. Discussion The electrical activity of the brain is sensitive to its oxygen supply. On high-altitude exposure, for example, EEG topographical changes are remarkable for an increase in delta and theta activities and a decrease in alpha activity (Papadelis et al., 2007). Furthermore, because EEG is profoundly affected by transient hypoxia (Ozaki et al., 1995) it can be considered as an indicator of cerebral metabolism (Jordan 1993). Hyperventilation is also known to cause, especially in children, an increase in synchronous EEG slow activity, together with a slowing of fast rhythms (Niedermeyer and Lopes da Silva 2000) and vasoconstriction of the cerebral resistance vessels (Visser et al., 1996b). However, the effects of oxygen
12
exposure on cerebral activity are not well known. Few studies have quantified by EEG the effects of O2 toxicity on cerebral activity (Visser et al., 1996a,b). They showed a global increase in power, which occurred in the first part of the experiment and reached a steady state after 100% O2 inhalation was started. An increase in high-frequency alpha activity and a non-correlated decrease in low frequency alpha were found after decompression. The gradual increase in low-frequency alpha activity toward the end of the experiment was not necessarily an effect of oxygen but instead might have been a slow effect of compression. Nonetheless, the results of this study do not allow to draw definitive conclusions, particularly in the subject who reported a seizure. Our findings, associated to a time-dependent decrease in fast delta activity in the posterior regions and an increase in alpha activity in the same regions, showed a behaviour opposite that of the hypoxia condition. Since delta activity is a sign that the cortical inhibitory mechanisms (e.g., the circuit of pyramidal and stellate cells) are more active, the delta decrease observed during oxygen breathing could represent reduced performance of cortical inhibitory mechanisms. During prolonged exposure to HBO in these professional divers, EEG changes mainly in the delta and alpha ranges, albeit with small latencies, were reproducible among the subjects. The hypothesis for vasomotor cerebral activity is very attractive. This phenomenon would consist of vasoconstriction followed by vasodilatation. The opposing behaviour (increase in delta activity and decrease in the alpha band) described by Manganotti et al. after transcranial magnetic stimulation (TMS) over the cerebral motor area (Manganotti et al., 2012) and the reduction in cerebral vasomotor reactivity after TMS stimulation observed by Vernieri et al. (2009) is consistent with this hypothesis. A pattern of vasoconstriction and vasodilatation during oxygen breathing previously reported in animals (Torbati et al., 1979) may be related to complex nitric oxide metabolism (Demchenko et al., 2000) and the role of oxygen free radicals (Elayan et al., 2000). Nitric oxide (NO) is a key intercellular
13
messenger in cerebral and peripheral hemodynamics. It plays a central role in regulating cerebral blood flow and cell viability and in protecting nerve cells or fibers against pathogenic factors associated with cerebral ischemia, trauma, and hemorrhage. Specifically, cerebral blood flow is increased and cerebral vascular resistance is decreased by NO derived from endothelial cells, autonomic nitrergic nerves, or brain neurons under resting and stimulated conditions (Toda et al., 2009). We would assume that during oxygen breathing vasoconstriction is followed by vasodilatation with an increase in alpha activity in the posterior regions, independently of oxygen pressure (Weaver and Howe, 1992). Vascular response to hyperoxia leads to a reduction in cerebral blood flow. In the present study, the hyperoxic conditions produced a residual effect on the CNS, with a significant alteration of both slow and fast rhythms until 8 min of the AIRpost stage. Neuron functional connectivity due to hyperoxia may be responsible for the observed increase in the relative alpha power in the posterior regions. Finally, we observed that the significant increase in alpha activity persisted until 20 minutes of EEG recording only in the posterior regions. Despite strict wakefulness maintenance while recording, we cannot completely exclude that a slight increase in theta components as a marker of increased drowsiness may have occurred. Nonetheless, our results showed a decrease in theta activity, especially during O2 breathing. While many mobile EEG systems have been developed in different contexts and for a variety of purposes (i.e., in epilepsy, sleep, brain computer interface) (Askamp and van Putten, 2013), here we focus on their potential application in hyperbaric chambers. Electroencephalographic studies in hyperbaric chamber are not common. The main limitations are the problems with recording the EEG signal owing to the fact that the chamber has a steel casing, i.e., a Faraday cage, and that the power supply of the equipment inside the chamber cannot be powered by alternating current because it would generate sparks. The development of Holter apparatuses, i.e., devices that amplify signals close to the subject, allows for using low-voltage direct current (DC) power supply, which does not generate dangerous sparks during circuit opening and closing. Furthermore, amplification near the signal source allows for the use of short
14
cables thus minimizing the eddy currents. With the advent of Bluetooth technology, cables have been replaced with low-length radio waves that allow its use in all types of hyperbaric chambers. The main advantage of Bluetooth technology is the improved EEG signal quality over older technologies, thus reducing environmental or cable movement interferences. Oxygen toxicity is a problem in diving, with potentially fatal consequences for the diver in the water. When divers use a closed-circuit oxygen rebreathing apparatus, they are breathing O2 100% and are exposed to hyperoxide conditions. In this scenario, the body could be affected in different ways depending on the type of exposure. Central nervous system toxicity is caused by short exposure to high partial pressures of oxygen at greater than atmospheric pressure. Pulmonary and ocular toxicity result from longer exposure to elevated oxygen levels at normal pressure. Exposure to HBO can cause oxidative damage to cell membranes, collapse of the pulmonary alveoli, retinal detachment, and seizures (Acott 1999; Beehler, 1964, Alcaraz-García et al., 2008). Symptoms may include disorientation, breathing problems, and vision changes such as myopia. Since symptoms of visual disturbance, ear problems, dizziness, confusion, and nausea can all arise from such common factors as narcosis, congestion and coldness, the diagnosis of CNS O2T in divers prior to seizure is difficult to establish. The findings of this study, i.e., significant changes in the relative power of both delta and alpha oscillations between oxygen exposure and the baseline stage, suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygen-toxicity diving-related problems”. The approach to monitoring cerebral power spectra could be useful to evaluate the physiological effects of oxygen breathing and to delineate different diving profiles, providing indications on the potential risk of a dive. With this study we have constructed a normative dataset of EEG responses to HBO, to which the EEGs of “divers experiencing problems" can be compared statistically. Training activity in hyperbaric chambers during EEG measurements can be very important in the selection of divers who are more suitable to deal with a stressful undersea environment. We are also planning in future studies to include EEG evaluation during oxygen breathing at a pressure of 1 ATA and air at 2.8 ATA in order to determine how oxygen and hyperbaric pressure variables affect
15
bioelectrical activity in human professional divers.
Conflict of interest The authors report no conflicts of interest.
Acknowledgements This study was funded by grant SMD L-023: Rilevazioni elettrofisiologiche in immersione; Cap. 1322, Italian Ministry of Defense, Direzione Generale della Sanità Militare, 2010.
References
Acott C. Oxygen Toxicity – A brief history of Oxygen in Diving. SPUMS J 1999; 29:150-155.
16
Alcaraz-García MJ, Albaladejo MD, Acevedo C, Olea A, Zamora S, Martínez P, et al. Effects of hyperoxia on biomarkers of oxidative stress in closed-circuit oxygen military divers. J Physiol Biochem 2008;6 4:135-41.
Arieli R, Shochat T, Adir Y. CNS toxicity in closed-circuit oxygen diving: symptoms reported from 2527 dives. Aviat Space Environ Med 2006; 77:526-32.
Babiloni C, Vecchio F, Lizio R, Ferri R, Rodriguez G, Marzano N, et al. Resting state cortical rhythms in mild cognitive impairment and Alzheimer's disease: electroencephalographic evidence. J Alzheimers Dis 2011; 26:201-14.
Beehler, CC. Oxygen and the eye. Survey of Ophthalmology 1964; 45: 549–60.
Bitterman N, Melamed Y, Perlman I. CNS oxygen toxicity in the rat: role of ambient illumination. Undersea Biomed Res 1986; 13:19-25.
Butler FK Jr, Knafelc ME. Screening for oxygen intolerance in U.S. Navy divers. Undersea Biomed res 1986; 13:91-98.
Butler FK Jr, Thalmann ED. Central nervous system oxygen toxicity in closed circuit scuba divers II. Undersea Biomed Res 1986; 13:193-223.
17
Cohn R, Gersch I. Changes in brain potentials during convulsions induced by oxygen under pressure. J Neurophysiol 1945; 8:155-60.
Demchenko IT, Boso AE, O’Neill TJ, Bennett PB, Piantadosi CA. Nitric oxide and cerebral blood flow responses to hyperbaric oxygen. J Appl Physiol 2000; 88:1381-9.
Donald KW. Oxygen poisoning in man. Br Med J 1947; 1:667-72.
Elayan IM, Axley MJ, Prasad PV, Ahlers ST, Auker CR. Effect of hyperbaric oxygen treatment on nitric oxide and oxygen free radicals in rat brain. J Neurophysiol 2000; 83:2022-9.
Formaggio E, Storti SF, Avesani M, Cerini R, Milanese F, Gasparini A, et al. EEG and fMRI coregistration to investigate the cortical oscillatory activities during finger movement. Brain Topogr 2008; 21:100-11.
Formaggio E, Storti SF, Boscolo Galazzo I, Gandolfi M, Geroin G, Smania N, et al. Modulation of eventrelated desynchronization in robot-assisted hand performance: brain oscillatory changes in active, passive and imagined movements. J Neuroeng Rehabi 2013; doi: 10.1186/1743-0003-10-24.
18
Harabin AL. Human central nervous system oxygen toxicity data from 1945 to 1986. Bethesda, MD: Naval Medical Research Institute 1993; (NMRI Report 93-03).
Ibner C, Ancoli-Israel S, Chesson A, Quan SF (eds). The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specifications. 1st edition. Westchester, III: American Academy of Sleep Medicine; 2007.
Imlay JA. Pathways of oxidative damage. Annu Rev Microbiol 2003; 57:395-418.
Jordan KG. Continuous EEG and evoked potential monitoring in the neuroscience intensive care unit. J Clin Neurophysiol 1993; 10:445-75.
Lemaire C, Rostain JC. The High Pressure Nervous Syndrome and Performance. Marseille: Octares, 1988.
Manganotti P, Formaggio E, Storti SF, De Massari D, Zamboni A, Bertoldo A, et al. Time-frequency analysis of short-lasting modulation of EEG induced by intracortical and transcallosal paired TMS over motor areas. J Neurophysiol 2012; 107:2475-84.
Morris LE, Knott JR, Pittinger CB. Electroencephalographic and blood gas analysis in human surgical patients during xenon anesthesia. Anesthesiology 1955; 16:312-9.
19
Murata M, Suzuki M, Hasegawa Y, Nohara S, Kurachi M. Improvement of occipital alpha activity by repetitive hyperbaric oxygen therapy in patients with carbon monoxide poisoning: a possible indicator for treatment efficacy. J Neurol Sci 2005; 235:69-74.
Niedermeyer E, Lopes da Silva FH. Electroencephalography: Basic Principles, Clinical Applications, and Related Fields. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 2000.
Neubauer RA. High dose oxygen therapy in stroke. Eur J Neurol 2000; 7:151.
Okuda S, Matsuoka S, Mohri M. Topographic electroencephalographic studies in a hyperbaric environment--specific reference to high pressure nervous syndrome. J UOEH. 1988;10:247-61.
Ozaki H, Watanabe S, Suzuki H. Topographic EEG changes due to hypobaric hypoxia at simulated high altitude. Electroencephalogr Clin Neurophysiol 1995; 94:349-56.
Papadelis C, Kourtidou-Papadeli C, Bamidis PD, Maglaveras N, Pappas K. The effect of hypobaric hypoxia on multichannel EEG signal complexity. Clin Neurophysiol 2007; 118:31-52.
20
Pastena L, Mainardi G, Faralli F, Gagliardi R. Analysis of cerebral bioelectrical activity during the compression phase of a saturation dive. Aviat Space Environ Med 1999; 70:270-6.
Pastena L, Faralli F, Mainardi G, Gagliardi R. EEG patterns associated with nitrogen narcosis (breathing air at 9 ATA). Aviat Space Environ Med 2005; 76:1031-6.
Piantadosi CA, Clinton RL, Thalmann ED. Prolonged oxygen exposures in immersed exercising divers at 25 fsw (1.76 ATA). Undersea Biomed Res 1979; 6:347-56.
Rostain JC, Charpy JP. Effects upon the EEG of psychometric performance during deep dives in heliumoxygen atmosphere. Electroencephalogr Clin Neurophysiol 1976; 40:571-84.
Storti SF, Formaggio E, Franchini E, Bongiovanni LG, Cerini R, Fiaschi A, et al. A multimodal imaging approach to the evaluation of post-traumatic epilepsy. Magn Reson Mater Phy 2012; 25:345-60.
Takahashi TOY, Okamoto K, Yoshimura Y, Matsuoka Y, Oomori N. The current status of carbon monoxide poisoning and treatment with hyperbaric oxygen. Rinsho Seishin Igaku 1998;27:701–6.
Toda N, Ayajiki K, Okamura T. Cerebral blood flow regulation by nitric oxide: recent advances. Pharmacol Rev 2009; 61:62-97.
21
Torbati D, Parolla D, Lavy S. Organ blood flow, cardiac output, arterial blood pressure and vascular resistance in rats exposed to various oxygen pressures. Aviat Space Environ Med 1979; 50:256-63.
Vernieri F, Maggio P, Tibuzzi F, Filippi MM, Pasqualetti P, Melgari JM, et al. High frequency repetitive transcranial magnetic stimulation decreases cerebral vasomotor reactivity. Clin Neurophysiol 2009; 120:1188-94.
Visser GH, van Hulst RA, Wieneke GH, van Huffelen AC. The contribution of conventional and quantitative electroencephalography during monitoring of exposure to hyperbaric oxygen. Undersea Hyperb Med 1996a; 23:91-8.
Visser GH, Van Hulst RA, Wieneke GH, Van Huffelen AC. Transcranial Doppler sonographic measurements of middle cerebral artery flow velocity during hyperbaric oxygen exposures. Undersea Hyperb Med 1996; 23:157-65.
Weaver LK, Hopkins RO, Chan KJ, Churchill S, Elliott CG, Clemmer TP, et al. Hyperbaric oxygen for acute carbon monoxide poisoning. N Engl J Med 2002; 347:1057-67.
22
Weaver LK, Howe S. Normobaric measurement of arterial oxygen tension in subjects exposed to hyperbaric oxygen. Chest 1992; 102:1175-81.
Zal’tsman GL. Physiological principles of sojourn of a human in conditions of raised pressure of gases medium [in Russian]. Ohio: Foreign Technology Division, Wright Patterson Air Force Base 1967; AD 655, 360.
23
Figure legends
Figure 1. (A) 32-channel HydroCel Geodesic Sensor Net Map. (B) A navy diver is wearing the EEG cap outside the hyperbaric chamber. (C) During the experiment, the subject is reclined on a cot with eyes closed. The 32-channel EEG system is connected via Bluetooth wireless transmission to a notebook that visualizes the EEG signal outside the hyperbaric chamber.
Figure 2. (A) Dive profile and recording sessions. The subject is reclined on a cot with eyes closed. A 20minute baseline EEG recording was made at 1 ATA breathing air (air) in the open hyperbaric chamber. In the closed chamber, a 2-min compression profile (descent rate 9 m min-1) breathing air was used to reach the next stage at a pressure of 2.8 ATA (compression). At this pressure, the subject breathed pure oxygen via an oronasal mask (oxygen) and a 20-min EEG was acquired. After decompression, back on air breathing, the EEG of each subject was recorded for 20 min, discarding the first 2 min (ascent rate 9 m min-1) (decompression). (B) Example of an EEG signal after preprocessing: air (AIRpre), end of oxygen breathing at minute 20 (O2), and end of air breathing after decompression at minute 20 (AIRpost).
Figure 3. Power spectral densities (μV2/Hz) of the EEG signal (only 19 channels visualized in the international 10-20 system positions) obtained by FFT on 2 s epochs, in one professional diver. (A) O2 min 18-20 (green) vs. AIRpre average of 20-minute EEG recording (blue). (B) AIRpost min 8 (red) vs. AIRpre average of 20-minute EEG recording (blue). The y-axis has a different range for each channel in order to capture the differences between the spectra. The scale is the same in A and B.
24
Figure 4. Relative EEG power maps in delta, theta, alpha, lower beta and upper beta bands averaged from 11 subjects. During air breathing averaged in 20 min (AIRpre), during O2 breathing (1 ATA) (O2) and during air breathing after decompression (AIRpost) at 1, 2, 5, 8, 10–12, 15–17, 18–20 min. Red coding indicates maximal relative power. At 1, 2, 5, 8 min after decompression there is a transient phase where pressure and oxygen effects take some time to stabilize (e.g., saturation effect).
Figure 5. T-maps O2 vs. AIRpre (A) and AIRpost vs. AIRpre (B) of relative EEG power in delta, theta, alpha, beta1 and beta2 bands. Each minute of EEG recording during O2 and AIRpost was compared to the entire AIRpre recording (average of 20 minutes), respectively. The threshold is p < 0.05 (|t| > 2.228) at 1, 8, 10–12 and 15–17 min (11 recordings) and p < 0.05 (|t| > 1.812) at 5 and 18–20 min (10 recordings). (+) indicates significance for Bonferroni correction. The effects in 'oxygen vs. air' and ‘after decompression vs. air’ show transients in opposite directions.
25
Table 1. Analysis of variance (ANOVA) for repeated measures applied to the relative power with the factor “condition” (AIRpre, O2, AIRpost).
ANOVA factor “condition” (AIRpre, O2, AIRpost) 1'
2'
5'
8'
10'-12'
15'-17'
18'-20'
F(2,20) = 20.397
F(2,18) = 18.422
F(2,18) = 14.933
F(2,20) = 15.605
F(2,20) = 7.131
F(2,20) = 6.064
F(2,18) = 4.897
p < 0.001
p < 0.001
p < 0.001
p < 0.001
p < 0.01
p < 0.01
p < 0.05
F(2,20) = 7.650
F(2,18) = 7.881
F(2,18) = 4.866
F(2,20) = 8.436
F(2,20) = 7.856
F(2,18) = 6.769
p < 0.005
p < 0.005
p < 0.01
F(2,20) = 3.931
F(2,20) = 5.259
δ
θ
n.s. p < 0.005
p < 0.005
p < 0.05
F(2,20) = 15.592
F(1.304, 11.732) = 16.725
F(2,18) = 9.217
p < 0.001
p < 0.005
p < 0.005
p < 0.01
p < 0.05
p < 0.05
n.s.
n.s.
n.s.
n.s.
n.s.
n.s.
F(2,18) = 4.629
F(2,20) = 3.737
F(2,20) = 4.569
α
β1
F(2,20) = 4.432
β2
F(2,20) = 7.043
n.s.
n.s. p < 0.05
n.s.
F(2,18) = 4.207 n.s.
p < 0.05
p < 0.05
p < 0.05
p < 0.05
n.s. = not significant
26
27
28
30
31
Abstract Objective: The aim was to investigate and define possible alterations in cerebral activity during prolonged hyperbaric oxygen exposure and decompression as compared to baseline activity. Methods: Thirty-two channel electroencephalography (EEG) was recorded with a Bluetooth EEG system in 11 subjects. A 20-minute EEG recording was carried out under three different conditions: breathing air inside a hyperbaric chamber at sea level; breathing oxygen at a simulated depth of 18 msw; breathing air at sea level after decompression. Relative EEG power was estimated in frequency ranges. Results: During oxygen breathing, brain activity showed an early fast delta decrease in the posterior regions, with a synchronous and significant increase in alpha in the same regions. After decompression, the delta relative power decrease was uniformly distributed over the cerebral cortex until minute 8, and the alpha relative power was maximal in the posterior regions during the first 2 minutes. Conclusions: These results may be relevant for establishing a reference point in future studies on oxygen-sensitive subjects who reported problems during oxygen diving. Significance: Significant changes in EEG relative power suggest that it may be possible to define and recognize landmarks of oxygen-induced brain activity, which would be useful in the medical treatment of subjects reporting “oxygen-toxicity diving-related problems".
32
