Sleep, 16:S76-S78
© 1993 American Sleep Disorden; Association and Sleep Research Society
The Utility of Topographic EEG Mapping in
Obstructive Sleep Apnea Syndrome
Joyce A. Walsleben, Edward B. O'Malley, Kenneth Bonnet,
Robert G. Norman and David M. Rapoport
*Division of Pulmonary and Critical Care Medicine and tDepartment of Psychiatry,
New York University/Bellevue Hospital Medical Center, New York, New York, U.S.A.
S~~mary: A study ~as undertaken to determine whether topographic brain mapping of standard EEG and cogmtIve evoked potentials would provide additional information for the detection of subtle brain abnormalities
associated with obstructive sleep apnea. During nocturnal sleep, significant brain activity was detected in frontal
re~o~s not usually. monitored with standard sleep recordings. Moreover, preliminary results suggest that total brain
activity decreased In association with apneic events and depth of O 2 desaturation. Also, component asymmetry in
the P300 waveform observed in brain areas not typically recorded improved with treatment. We conclude that the
use of topographic mapping adds important information to the study of brain function during sleep and sleep apnea.
Key Words: Obstructive sleep apnea- Topographic brain mapping-Evoked potentials-EEG-Hypoxia.
Obstructive sleep apnea syndrome (OSA) is characterized by abnormalities in sleep architecture, recurrent hypoxia during sleep and deficits in daytime
cognitive abilities. Previous studies using electrophysiologic measures, including evoked potentials (EPs),
have detected subtle changes in brain function associated with OSA (1-3). However, it is not clear how
these changes in brain function are related to the repetitive hypoxia and/or sleep fragmentation associated
with OSA. We have begun to examine these issues
utilizing multichannel topographic brain electrical activity mapping (TBM).
TBM extends the clinical utility of standard electroencephalographic (EEG) recordings by providing spectral analyses that can be displayed using spatial mapping and temporal summation of extended EEG epochs.
EPs can also be analyzed using topographic mapping
to evaluate waveform symmetry and the interaction
of multiple components previously not appreciable by
eye (4,5). To date, few studies have applied these techniques to sleep. Brain maps of sleep stages have been
shown to be stable and reproducible during normal
daytime naps (6), and some EEG abnormalities have
been noted during apneas in patients with OSA (7,8).
This paper describes preliminary findings utilizing
topographic mapping during nocturnal sleep in three
patients with OSA and four age-matched normal controls.
METHODS
Twenty-one channels of EEG (International 10-20
System), including the standard polysomnographic
montage, were recorded during diagnostic sleep studies
and during initiation of continuous positive airway
pressure (CPAP) therapy on a separate night. All channels were referenced to linked ears. Respiratory and
EEG data were simultaneously collected on the Biologic Sleep Scan system and the Bio-logic Brain Atlas
III system and processed off-line. Care was taken to
time-lock the two systems using the recognition of
common waveform events. Multiple sections of interest were isolated, screened for artifact and subjected
to fast Fourier transform (FFT) to provide power spectral arrays or maps, which were then compared.
In a separate series of experiments, standard auditory P300 paradigms were topographically mapped
during morning and afternoon wake time sessions. Single trials were screened for artifact and morphology
and analyzed by a NeuroSciences brain mapping system. Trials were collapsed to provide averaged P300
data. The developing EP waveforms were displayed
with 2-second sequential maps.
RESULTS
Preliminary results based on the seven subjects studied to date indicate the following:
S76
','
,.
TOPOGRAPHIC EEG M APPING I N OSA
S77
...,
4-7.5
8-12
12.5-24
(C)
B.nd. !Hit
4-7.5
8-12
12.5-24
-..
FIG. I. Toposraphlc brain maps (TOMs) utili7.lng data from 21 scalp electrodes (International 10-20 System) oollected with a Bio-Ioglc
Dr.lIn Alias 111 system dunng nocturnal sleep In normals and apnelcs. A) TBMs ~re genefllted from representauve JO·second epochs of
stage 2 sleep dunng ('arl)' and late portions or the nighl in a normal and an apneiC. Note the Similar pattern and level oraclivll), aCf1m
the night "'lIhln each subject. 8) TIlMs (14 seconds) berore. during and after an apnea. There 15 a decrease: In aCllvlly that ooincidcs With
the apneic even!. C) TOMs dunn& normal slase 2 sleep and apneic events. Note that bram aetlvlty diminishes With decreasmg 0 , satuflIuon
and Increa~m& e\cnt durallon
•
r
r
[
[
I
I. Consislcn l wi lh data from daytime nap studies delayed latency of all components. This was most ev(5). we found lopographic mapping ofEEG during noc- iden t in the right parietal area pretreatment and im·
turnal slccp dCTllonstralCd a reproducible pattern of proved with 2 weeks of CPAP treatment (Fig. 2).
brain activity in various sleep Slages across the night,
with in and between su bjects. T hi s sta bi lity formed the
basis from which we compared changes in brain acDI SCUSSION
tivity during apneic events (Fig. IA).
2. The topographic map of the EEG during sleep
T he reproducibility of brai n maps in normals and
provided add itional information. Alpha and theta ae- a pneics during stage 2 sleep suggests that TUM does
tivity frequently oceurred in frontal areas out of range provide a stable representation of brain electrical acof standard electrode sites used for sleep scoring (Fig. tivity during sleep. TUM also capt ures additional in IA).
formation. Specifically. activity in frontal brain areas
3. Although the topographic distribution of activity from all frequency bands is often u nd et ~ t cd by stanwithin each range of freque ncies examined remained dard slecp el~ trode placements. We have observed
simila r to that present during stable non rapid eye multiple short, poten tially disrupting, arousal-like
movement (N REM ) stage 2 sleep. a consistent d~reasc events occurring in these areas. This information may
in power occurred in association with all apneic events. be relevant to the analysis of associations between rePower returned to baseline following single apneas (Fig. spiratory events. arousal and cognitive defi ci ts.
10).
Our observat ion of an association between the mag4. As apneic events increased in frequency , duration ni tude of the decreased brain activity during apneic
and oxygen desaturation . the decrease in brain activity even ts and the dura tion and magnitude of 0 1 desaturaduring each event became more prominent (Fig. I C). tion suggests that decreased brain activity may result
5. During daytime EP testing in one patient wit h from increasing hypoxia. Alternatively, the change in
apnea . topographic maps of the P300 waveform re- activity ma y represent 11 change in brain state that
vea led sign ificant asymmetry of NIOO and severely precipita tes the apneic even t and/or influences its duSIN'P. Vol. 16. No.8. 199J
578
1. A. WALSLEBEN ETAL.
PRE - CPAP TREATMENT
POST- CPAP TREATMENT
,...,
0 .0
"' IG . 2. Topographic mapping of the P300 waveform before (PRE) and aner (I'()Sn Z "'cds of CI'AP treatment In an apneic pallent.
DaHl were collccll-d and anaJ)'fcd o n a NeuroSciences Bram Mappin g system. Maps rcprt'5(:nl conlLguo us 2-sccond displays. Markers I
and 2: In the PRE condnlon mdH::lIc the SIan and peak of NIOO, rcSpe("11Vely. The marker In the POST condnion Indicates the peak of
the NIOO. Note the decreased lalency and Improved symmelI')' orN 100 In the POST condillon
ration. Morc detailed analysis is necessary to differen tiate between Ihcsc possibilities.
Additional information is also available from to·
pographic mapping of EPs. Standard EP paradigms
collapse all trial s into onc latency average for three
central elect rode sites. This approach would have
missed the 'lbnormal asymmetry and underestimated
co mponent latencies. The N I 00 asymmetry is panic·
ularly intriguing because it has been shown that there
is a si milar asymmetry associated with parietal areas
sensitive to metabolic encephalopathy (9). Furthermore, the observed cha nges in P300 waveform morphology :Uld latency occurred in brain areas subserving
the neuropsychological substrates of sustained and spatial attention , functions known to be deficient in patients with OSA.
In conclusion, our preliminary findings suggest that
topographic mapping should be added to the pool of
elcctroph ysiological techniques used to examine neu·
rophysiology and brain fun ction during sleep and ob·
structive sleep apnea.
REFERENCES
Walsleben JA. SqUlfCS NK. Rothenberger V. Audnory e\ent·
rehued potenllats and brnln dysfuneuon In steep apnea. £/«.
{rQ('nU.'flilaloKr elm Nl'lIrcphl'$lo/1989;74:297-3 11.
SIN'P. Vol. 16. NQ, 8. 199J
2, Rumbaek L.. Krieger J. Kum D Audnory event -related potenlIals In obslfueuve sleep apnea: elfeel5 of treatment wi th nasal
conh nuo us petSlhve auways pressure EI«trOl'nct'phulogr Cbn
Nl'unJph),sIOII99t;80: 454-7.
3. Edwards DA, IJearpa rk 11M. ('ostas L, Penlgrew AG, HendersonSman DJ. Sulhvan ('E. Bralnstem auditory evoked responses In
adults With moderate to snere obslrucllve sleep apnea before
and atler treatment. SIN'" Rl'S 1989: 18:228.
4 J)ulfy Fll. Mau re r K. Estabhshment ofguidelincs for the usc of
topographic mapping In ctinlcal neuroph}'siolog). In : Maurer K,
cd. Topogral"uc mapping of EE(; and l'I"Oke(/ poIen/lIlls, IJerhn.
Heidelberg, New York: Spnnger. 1989.
5. Duffy Fl! . il Tlun elcc trical aehvlly mapping: clmical applications.
l'sJ'f'hllJlry RI's 1989;29:379-84.
6. Buchsbaum MS. Mendelson WIt Dunca n WC, Coppola R. Kelsoe J. Gil hn Jc, Topogra phic conical mapping of EEG slee p
stages during da ytim e naps in no rmal su bJcc ts. SI"ep 1982:5:248-
55.
7. Sagalcs T . Sampol G, G imeno v. et al. Brain EEG topograph)
In patients With chronic obslrucli\e pulmonary disease and obstructive slee p apnea syndrome. Sleep Res 199 1:20A:317.
8_ 5\ anbo'1 E, Gulllcmmault C Topoglllphic EEG analysis during
obstructive sleep apncas Brain Topography 1990;2:309.
9. Mesulam M-M AttentIon, confusional states and negleet In
MC5ulam M-M. cd. l'rmC/plt's o!brharioral ,,('Urology. Phlladel·
phla: F. A J)a\IS Company, 1985:125-68