Contrainte Psycho-Physiques et Electrophysiologiques sur le codage de la stimulation électrique chez les sujets porteurs d'un implant cochléaire

IV/ PEAEP et psychophysique

Les études précédentes de ce chapitre ont montré la possibilité d'obtenir des PEAEP et de les caractériser en fonction de plusieurs paramètres de la stimulation. D'après les résultats décrits, chacun de ces paramètres varie de manière spécifique au sujet testé.

Des études contenues précédentes mettent aussi l'accent sur l'importance de la spécificité des contraintes psychophysiques des sujets testés pour adapter le signal transmis via l'implant cochléaire au système auditif.

L'objectif des travaux qui suivront sera d'évaluer la possibilité de mesurer de manière objective, avec le recueil et l'analyse des PEAEP, les contraintes psychophysiques du patient testé, cela afin de développer des techniques objectives lors du réglage qui ne font pas intervenir le patient.

a/ Apparition des PEAEP et seuils de détection

Comme les mesures psychophysiques n'ont pas montré de différences significatives des seuils de
détection en fonction de la fréquence de stimulation de 50 à 1.000 Hz, non avons étudier la
comparaison entre le seuil de détection à une fréquence de stimulation de 300 Hz et l'apparition des

PEAEP ayant une fréquence de stimulation moindre (60 Hz). Des études préliminaires (Gallego et al, 1996, 1997) ont montré sur une population réduite la possibilité d'estimer les seuils de détection à partir des PEAEP.

- Article 17 :

CORRELATION BETWEEN ELECTRICAL AUDITORY BRAINSTEM RESPONSE AND
PERCEPTUAL THRESHOLDS IN DIGISONIC COCHLEAR IMPLANT USERS

E. Truy, S. Gallego, J.M. Chanel, L. Collet, A. Morgon

Laryngoscope, 1998, 108, 554-559

L'étude des PEAEP en fonction de l'intensité de stimulation montre qu'il est possible d'obtenir des PEAEP à faible niveau de sonie. Nous avons comparé sur une population de 9 sujets porteurs de l'implant cochléaire Digisonic l'apparition des PEAEP (pour une stimulation de fréquence 60 Hz) avec les seuils de perception subjectifs à la fréquence standard de 300 Hz (fréquence de réglage). Nous avons effectué pour chaque électrode testée un seuil sur au moins 12 niveaux d'intensité de stimulation différents (de 12 à 32, la plupart du temps 16). Pour chaque sujet, nous avons testé 3 à 4 électrodes équiréparties. Au total 31 électrodes ont été testées. Une comparaison sur les 31 électrodes entre le seuil de détection et l'apparition des PEAEP montre une corrélation statistiquement significative p<0.001 avec R=0.98 (l'unité est en ps). La relation qui lie le seuil de perception (300 Hz) avec l'apparition des PEAEP (60Hz) explique 96 % de la variance (R²). On peut modéliser cette corrélation par une fonction linéaire telle que Seuil_Détection = 0.22 + 1.06 x Apparition_PEAEP. Le seuil d'apparition des PEAEP survient donc en moyenne au même niveau ou a un niveau légèrement supérieur au niveau de perception subjectif.

The Laryngoscope

Lippincott--Raven Publishers, Philadelphia (c) 1998 The American Laryngological, Rhinological and Otological Society, Inc.

Correlation Between Electrical Auditory

Brainstem Response and Perceptual

Thresholds in Digisonic Cochlear

Implant Users

Eric Truy, MD; Stéphane Gallego, MSc; Jean-Marc Chanal, MSc; Lionel Collet, MD, PhD; Alain Morgon, MD

Objectives: To examine the relationships between psychophysical perceptions and the electrically evoked auditory brainstem responses (EABRs) in multichannel cochlear implant (CI) users and to determine the effectiveness of EABRs in electrode failure. Design: A descriptive study reported the EABR characteristics while the different electrodes were activated. Characteristics of the EABR and of the perceptual measures served as compared variables in a correlational study. Setting: The study was carried out in the audiology clinic of an otolaryngology department at a university hospital. Patients: The subjects consisted of vine consecutively selected habitual Digisonic DX1OR multichannel CI users. Seven patients were postlinguistically deafened adult patients; two were congenitally deaif children. Main Outcome Measures: Ipsilateral recordings were performed using a previously published method. Morphology, latency, and amplitude measures of the EABR recordings were described, computed, and compared with the literature data for EABRs obtained while activating other types of CI and for acoustically evoked ABRs. Correlations between EABRs and behavioral perception thresholds were analyzed using the parametric Pearson's correlation test. Results: EABRs allowed the authors to detect failure of no. 10 electrode integrity in one child. Perceptual thn-shold measures were fo und to be highly significa iy related to the EABR threshold across subjects and electrode position (n = 31, r = 0.98; P < 0.001; linear regression equation: perceptual threshold = 1.06 EABR threshold + 0.76). The latencies and amplitudes were found to be similar to those described in the literature. Conclusions: EABRs may be used to estimate settings

From the Department of Otorhinolaryngology (E.T., J-M.C, L.C., A.M.), de Chirurgie Cervico-Faciale et de Phoniatrie, Hôpital Edouard Herriot, Lyon, ?rance; and the Laboratoire "Perception et Mécanismes Auditifs" (E.T., S.G., J-nc., L.c.), Lyon, France.

Editor's Note: This Manuscript was accepted for publication June 9,

1997.

Laryngoscope 108
· April 1998 Truy et al.: Cochlear Imprants

Send Reprint Requests to Eric Truy, MD, Professor, Département d'Oto-Rhino-Laryngologie, de Chirurgie Cervico-Faciale et de Phoniatrie, Hôpital Edouard Herriot, Place d'Arsonval, 69437 Lyon, Cedex 03, France. for the Digisonic DX10 CI even in a pediatric population, although they cannot entirely replace behavioral measurements, especially in children. The EABR can be employed for electrode dysfunction diagnosis. Further studies are needed to determine whether recordings of EABR quality could contribute to the evaluation of functional prognosis during the rehabilitation. Key Words: Cochlear implant, electrically elicited auditory brainstem responses, human, objective measures, tuning aids.

Laryngoscope, 108:554-559, 1998

INTRODUCTION

Cochlear implantation has proven highly effective in adults and children, when suitable candidates are selected. The success of a cochlear implantation may be related to the degree of neural survival. It has been demonstrated that the major factor influencing neural survival is the duration of the deafness.^1,2 Studies have demonstrated that there is a correlation between the number of surviving spiral ganglion tells and the quality of the electrically evoked auditory brainstem responses (EABRs) in the deafened cat.³

Preoperatively, effective assessment is needed to estimate the potential benefit. Different tests are required to indicate or counterindicate cochlear implantation in a given candidate. Electrical tests have been proposed to estimate neural survival. The promontory electrical stimulation test is crude: it consists in eliciting psychoacoustic responses in alert adults, but it cannot be used in most children. Its results are not reliably predictive of postoperative performances in cases of long-term deafness in postlinguistically deafened adults. The use of EABR,⁴ of electrically evoked middle latency responses (EMLRs),⁵-⁷ and of electrically evoked stapedius reflex⁸ has been reported. Objective mea- sures further provide information according to ear selection and device selection.

Postoperatively, the stimulus artifact can be recorded, as a means of detecting electrode failure. Measurement of the average electrode voltage has been pro-

posed to improve stimulus artifact recording.⁹ EABR can be used to evaluate cochlear implant (CI) functioning. 10,11 The relationship between postoperative behavioral and preoperative objective measures is not strongly significant on statistical analysis: the objective thresholds are higher than the perceptual ones.¹¹ Postoperatively, condition, device fitting, and rehabilitation are very important stages that must be optimized to get the best results. Thus all potential ways of better adapting the implant's electrical to the subject's surviving individual electrophysiologic properties are very important to consider. The C-level (comfort threshold) and the T-level (threshold of perception) need to be determined Determining thresholds in young children is difficult; to date, especially in pediatric populations, only behavioral perceptual measures are used. Therefore EABR has also been proposed as a tool for device-fitting procedures. The presence of an EABR indicates not only that the device is functioning, but also that the patient is receiving auditory information. EABRs obtained from subjects implanted with various types of intracochlear multichannel devices have been reported. ^4,12-14 To be able to consider EABRs as a new tool for estimating CI thresholds, we have to demonstrate that relationship between behavioral perceptual measures and EABR threshold. The postoperative electrically evoked stapedius reflex threshold has been demonstrated to cor- relate with the T-level threshold.¹⁵ Cortical potentials (P300, topographic brain mapping, and mismatch negativity) can be recorded during CI functioning but to date have been proposed as a research tool only in populations, not in individuals.

The questions to be answered in our study were the following: 1. Can EABRs be reliably recorded through the Digisonic DX10 CI? and 2. Can the EABR provide information helpful in fitting the device in adults and children?

PATIENTS AND METHODS Patients

Seven postlinguistically deafened adults and two congenitally deaf children participated in the investigation. All were Digisonic DX10 CI (MXM Laboratories; Antibes, France) users. The ages of the seven adults ranged from 30 to 69 years at the date of implantation, and the two children were ages 4 and 5 years at the date of implantation. The etiologies were variable, including sudden deafness (n 1), trauma (n = 2), progressive genetic degenerative deafness (n = 1), otosclerosis (n = 1), progressive deafness of unknown origin (n = 2), and congenital deafness of unknown origin (in the two children). The duration of the profound or total deafness in the adults ranged from 2 to 5 years. Experience with the CI ranged from a minimum of 12 months to a maximum of 2 years and 6 months. All the patients but two had a total insertion (15 active electrodes with the Digisonic DX10 device); the two patients with partial insertion (patients B.RI. and J.B.F.) having, respectively, two and four active electrodes.

Methods

Laryngoscope 108: April 1998 Truy et al.: Cochlear Implants

Behavioral perceptual threshold recordings. Behavioral measures were obtained using a standard procedure in which the electrodes are stimulated by electrical impulses delivered through the individual wearable speech processor. Perceptual measures (T-level, C-level) were obtained for all active electrodes, the stimulation frequency rate being 300 Hz. Responses were recorded in the arbitrary device programming units. Postlinguistically deafened, and judged intelligent and cooperative, all adult patients were considered reliable according to their auditory percepts. In the children, thresholds were determined by a derived visual reinforcement audiometry technique.

EABR recordings. EABR recordings and behavioral measurement were performed for each subject in a single session.

The techniques of stimulation and of EABR recording have been previously described¹⁶; we will briefly report it again. The electrical stimulus was generated using a manufacturer-built interface device (Digistim system, MXM Laboratories; Antibes, France) connected to a personal computer via a serial port. This system served to trigger the evoked potential measurement system (Nicolet Pathfinder II).

The programming setup uses arbitrary units rather than pulse duration units to represent the intensity levels delivered. The stimulus frequency rate is 60 Hz. Since the same processor was used in all test conditions across subjects, its arbitrary units reliably represent stimulus level.

Gold disk electrodes were placed on the ipsilateral earlobe (Al position for a left ear, and A2 position for a right ear) and forehead (Fpz position) of the subject. This placement enabled ipsilateral recording.

In adults, EABRs were recorded in a quiet room with the patient lying supine. In the two children, EABRs were recorded after light barbiturate-induced sedation (Nembutal 2 mg/kg body weight, intrarectally).

The recording parameters were as follows: three averagings of 256 sweeps, 100-jaV sensitivity, analog band-pass filtering from 0.2 to 8000 Hz, 10-millisecond analysis time, 521-point window, and 50-kHz sampling frequency. The sensitivity of the Nicolet Pathfinder II was set to a 100-11V threshold. The first 400 microseconds of the EABR signal were excluded, to eliminate stimulus artifact. The signal was then filtered with a digital band-pass of 300 to 3000 Hz. The recording duration was 10 milliseconds per sweep. The 3 x 3 intercorrelation matrix was then computed to eliminate the most exceptional curves. The remaining curves were summed and averaged.

The first recording was made at perceptual comfort level; recording was repeated with step-by-step decreases in current intensity until the response disappeared.

The EABRs were analyzed for waves II, III, and V identification, EABR threshold, and waves II, III, and V amplitudes. Wave identification employed strong criteria. The analysis was performed by an independent electrophysiologist, using the gen-

1 2 3 4

Fig. 1. Two exemples of electronically evoked auditory brainstem response (EABR) recordings in two different patients. Good intraindividual reproducibility is evident, as are the possible variations of amplitudes and latencies of waves II, III, and V.

eral morphologie criteria for ABR wave identification proposed by Picton et al.¹⁷ These criteria were adapted to the peculiarities of electrically evoked ABRs, as described in the literature: the electrically elicited response occurs approximately 2 milliseconds earlier than with acoustic stimulation. We included only reproducible waves on each of two recordings.

Statistical Analysis

The possible relationship between the behavioral threshold and the EABR threshold was examined using a parametric Pearson's correlation test (P-to-reject-correlation 0.05): the higher the r value, the better the correlation.

RESULTS

Morphology, Amplitudes, and Latencies

EABRs were able to be recorded from all the subjects. Figure 1 shows two examples of EABR in two different patients. These two examples demonstrate the good quality of the curves obtained and the good intraindividual reproducibility on test-retest evaluation (the second session being performed 6 weeks later). Wave latency and amplitude could differ across subjects. It is noteworthy that the performance of these two subi ects had changed markedly after 6 months' rehabilitation. (Open-set word recognition was 44% in patient M.B. and 22% in patient B.Ri., with CI alone, without the aid of lip reading.)

Behavioral and EABR measurements could be obtained from 55 of a total of 56 electrodes. In patient A.M., we were able to obtain neither EABR (Fig. 2) nor behavioral measures for electrode no. 10; we concluded that there was failure of this electrode.

Figure 3 shows EABR recordings in the child (patient A.M.) stimulus intensity (arbitrary device units) decreased from level 1 to level 8. These curves were reproducible for each level of stimulus current intensity; the higher the intensity, the greater the various wave amplitudes. Three peaks are well identified: the first is that of w ie II (at about 1.25 milliseconds), the second of wave III (at about 2 milliseconds), and the third of wave V (at about 3.8 milliseconds).

0 1 2 3 4 5 6 7 8

Fig. 2. EABR recordings in child patient. Among the eight tested electrodes, we could not obtain any response in only one case (no. 10). The stimulus artifact has been eliminated from the various recordings.

TABLE I.
Mean Latency Values and Mean Amplitude Values.

	Means (ms) (DS)		Amplitudes (gV) (DS)
Wave II	1.28	(0.18)	0.24	(0.10)
Wave III	2.06	(0.19)	0.43	(0.16)
Wave V	3.90	(0.27)	0.37	(0.13)

Table I reports mean amplitude and latency for the various waves; the intervals between the waves are also reported.

Thresholds

Thresholds were able to be determined for all patients. Figures 4 and 5 show two examples. We looked for a possible correlation between objective threshold (determined by EABR) and subjective T-level threshold (determined by perceptual measurement). It is worth emphasizing once again the great difficulty of determining behavioral T- and C-level thresholds in pediatric patients soon after implantation. Subjective (ST) and objective (OT) threshold values were expressed in arbitrary device units, and then compared by Pearson's correlation testing. The correlation coefficient found was excellent (n = 31, r = 0.98; P < 0.001). The scatterplot illustrating this strongly positive linear relationship is presented in Figure 6. The simple regression equation (OT = 1.06 x ST + 0.22) further attests to the strength of the relationship. This also means that the objective thresholds underestimate perceptual thresholds, although the corrective factor is weak; the regression equation approximates y = X.

DISCUSSION

These results demonstrate the possibility of goodquality EABR recordings in Digisonic DX10 CI users, both adults and children.

EABR Compared With ABR

Our results agree with numerous previous published data showing earlier latencies with EABR than

0 1 2 3 4 5 6 7 8 9 10

IChildren Threshold : Patient MOI

Fig. 3. EABR in child patient. Eight levels of stimulation are shown for electrode no. 7, decreasing from level 1 to level 8.

Fig. 4. EABR recordings in adult patient. Thirty-two decreasing levels of stimulation are shown for electrode no. 10. The EABR threshold is identified where waves III and V vanish simultaneously (level 25).

with ABR.¹⁸-²⁰ This is easily explained by the direct stimulation of the acoustic nerve without delay attributable to travel through the outer and the middle ear. With acoustic stimulation in nondeaf patients, wave II is recorded with a mean 2.90-millisecond (#177;0.22) delay, wave III with a mean 3.84-millisecond (#177;0.20) delay, and wave V with a mean 5.60-millisecond (#177;0.20) delay²¹; other studies have reported quite similar delays: respectively, 2.92, 3.95, and 5.85 milliseconds for waves II, III, and V.²² In our experiment with EABR, wave II was observed with a mean 1.20-millisecond (#177;0.14) delay, wave III with a mean 1.89-millisecond (#177;0.20) delay, and wave V with a mean 4.09-millisecond (#177;0.26) delay. Interwave delays were identical to those recorded with acoustical stimulation in non-deaf subjects: wave II to wave III delay is 1.03 milliseconds, wave II to wave V delay is 2.81 milliseconds, and wave III to wave V delay is 1.89 mil

TABLE II.
Mean Values of Interwave Intervals.

Delays (ms)

Wave II-wave III 0.78

Wave II-wave V 2.61

Wave III-wave V 1.84

Laryngoscope 108: April 1998 Truy et al.: Cochleadments

liseconds with acoustic stimulation²² and were, respectively, recorded with 0.78-, 2.61-, and 1.84-millisecond latencies in our study. Table I presents these data.

In our patients the disappearance of waves III and V (defining objective threshold) was concomitant and without prior increase in the latencies of these waves, as mentioned. With acoustic stimulation, wave III disappears before wave V and latencies increase, with an updown procedure.

Tuning Aid Procedures

EABR can be affected by factors such as number of sweeps and subject state. The maps used in this study were obtained using standard clinical techniques with routine stimulation parameters and were the actual maps used by the subjects in their everyday rehabilitation program.

The correlations between subjective and objective thresholds were strong. The present data are the first published on Digisonic patients; other data have essentially concerned the Nucleus device.^23,24 These studies reported good but not perfect correlation coefficients (r value of 0.89 in Hodges et a1.²⁴); Shallop et al.²⁵ found EABR thresholds to approximate C-levels, although with several instances where EABR threshold exceeded Clevel by more than 20 device units. Brown et al.¹⁰ found that the majority of EABR thresholds fell between 30%

Fig. 5. EABR recordings in adult patient. Thirty-two decreasing levels of stimulation are shown for electrode no. 14. The EABR threshold is identified where waves III and V vanish simultaneously (level 29).

and 80% of dynamic range, and that only four EABR thresholds of a total of 115 exceeded C-level, and in no case by more than five device units. Mason et al.¹¹ found EABR thresholds to exceed T-level by an average of 35 units; results reported by Brown et al.¹⁰ are consistent

0 10

Perceptual Threshold

20 30

(equipment--determined units)

ABR Th.= 0.22 + 1.06 Per.Th. N = 31

R = 0.98 P< 0.001

·

· -
·

!je _
·

· · ,ff°

..- · · ·

. ·

E 35 t

'e 30

o
D-
o

c
c

_E. _

a

· c

*_c* 20

E

o

0

co

25

15 10 5

Fig. 6. Scatterplot and linear regression equation showing the relationship between behavioral thresholds (BTs) and EABR thresholds (objective thresholds [OTs]). Symbols represent the data obtained from seven adult cochlear implant users. The continuous line is the linear regression curve, and the dotted line plots the y = x equation.

with these data. In these two studies,¹°^,11 it is reported that on average the difference between EABR threshold and T-level is somewhat greater for adults than for children, because T-levels found in children may be slightly above threshold. More data are needed, from a larger pediatric Digisonic-implanted population, to compare adult and pediatric EABR thresholds. We did not record any preoperative EABRs to be compared with postoperative EABRs, as reported in some studies. Mason et al.¹¹ described preoperative EABR in children, but linear regression analysis revealed no strong correlation between preoperative EABR threshold and postoperative behavioral threshold measurements: objective threshold overestimated subjective threshold levels.

Objective threshold measurement is obviously of major importance in pediatric populations. We insist on the fact that, in the procedure described, the stimuli employed to evoke the EABRs are strictly the same as those delivered in auditory conditions during rehabilitation. This is one of the major differences with EABRs obtained through the Nucleus device'¹; EABR stimulation modes described include bipolar mode (BP) +5, +10 or +20, which are very different from those of normal implant functioning (BP or BP + 1). Abbas and Brown²⁶ reported a good correlation between subjective and objective thresholds using BP + 5 strategy in Nucleus CI users and in Ineraid CI users.

Electrode Dysfunction Diagnosis

In our sample we only obtained no artifact or EABR in the case of one electrode. This electrode dysfunction was confirmed by the absence of any signal when we used the stimulogram method, as described by other authors,²⁷ and the average electrode voltage method.²⁸ These two techniques are easy to perform and recordings are short, that no sedation is needed in adults or children. They can be easily employed during surgery to assess implant functioning. The average electrode voltage is a refined artifact recording method.

CONCLUSION

Different procedures are employed or assessed to evaluate the electrophysiologic responses evoked by a CI. We advocate the rapid development of objective methods in pediatric populations, which may provide useful information for tuning in children.

BIBLIOGRAPHY

1. Moore D. Postnatal development of the mammalian central auditory system and the neural consequences of auditory de- privation. Acta Otolaryngol Suppl (Stockh) 1983;421:19-30.

2. Webster DB, Webster M. Effect of neonatal conductive hearing loss on brain stem auditory nuclei. Ann Otol Rhinol Laryngol 1979;88:684-98.

3. Hall RD. Estimation of surviving spiral ganglion cells in the deaf cat using the electrically evoked auditory brainstem responses. Hear Res 1990;45:123-36.

4. Lambert PR, Ruth RA, Hodges AV. Multichannel cochlear implant and electrically evoked auditory brainstem responses in a child with labyrinthitis ossificans. Laryngoscope 1991;101:14-9.

5. Burton MJ, Miller JM, Kileny PR. Middle latency responses. Part I. Electrical and acoustic stimulation. Arch Otolaryngol Head Neck Surg 1989;115:432-457.

6. Burton MJ, Miller JM, Kileny PR. Middle latency responses; Part II. Variations among stimulation sites. Arch Otolaryngol Head Neck Surg 1989;115:458-63.

7. Truy E, Morgon A, Collet L, et al. Is round-window electrical test possible in children? Ado Otolaryngol 1993;48:114-9.

8. Jerger JF, Jenkins H, Fifer R, Mecklenburg D. Stapedius reflex to electrical stimulation in a patient with cochlear implant. Ann Otol Rhinol Laryngol 1989;95:151-7.

9. Heller JW, Sinopoli T, Fowler-Brehm N, Shallop JK. The characterization of averaged electrode voltages from the Nucleus cochlear implant. Proc Ann Int Conf IEEE 1992;13:1902-8.

10. Brown CJ, Abbas PJ, Fryauf-Bertschy H, Kelsay D, Gantz BJ. Intraoperative and postoperative electrically evoked auditory brainstem responses in nucleus cochlear implant users: implications for the fitting process. Ear Hear 1994; 15(2):168-76.

11. Mason SM, Sheppard S, Garnham CW, Lutman ME, O'Donoghue GM, Gibbin KP. Improving the relationship of intraoperative EABR thresholds to T-level in young children receiving the Nucleus cochlear implant. In:

Laryngoscope 108: April 1998

Truy et al.: Cochlear Implants

Hochmair-Desoyer IJ, Hochmair ES, eds. Advances in Cochlear Implants. Wien: Manz; 1993:44-9.

12. Allum JH, Shallop JK, Hotz M, Pfaltz CR. Characteristics of electrically evoked auditory brainstem responses elicited with the Nucleus 22-electrode intracochlear implant. Scand Audiol 1990;19:263-7.

13. Gardi JN. Human brain stem and middle latency responses to electrical stimulation: preliminary observations. In: Schindler RA, Mezernich MM, eds. Cochlear implants. New York: Raven; 1985:351-63.

14. Shallop JK, Beiter AL, Goin DW, Mischke RE. Electrically evoked auditory brainstem responses (EABR) and middle latency responses (MLR) obtained from patients with the Nucleus multichannel cochlear implant. Ear Hear 1990; 11(1):5-15.

15. Hodges AV. The relationship between electric auditory evoked responses and psychophysical percepts obtained through a Nucleus 22-channel cochlear implant. Charlottesville, VA: University of Virginia, 1990. Dissertation.

16. Gallego S, Micheyl C, Berger-Vachon C, Truy E, Morgon A, Collet L. Ipsilateral ABR with cochlear implant. Acta Otolaryngol (Stockh) 1996;116:1604-10.

17. Picton TW, Hillyard SA, Krausz HI, Gallambos R. Human auditory evoked potentials. Part I. Evaluation of components. Electroencephalogr Clin Neurophysiol 1974;36:179-190.

18. Pelizzone m, Kasper A, Montandon P. Electrically evoked responses in cochlear implant patients. Audiology 1989;28: 230-8.

19. Van Den Honert C, Stypulkowsky PH. Characterization of the electrically evoked auditory brainstem response (ABR) in cats and in humans. Hear Res 1986;21:109-26.

20. Abbas PJ, Brown CJ. Electrically evoked brainstem potentials in cochlear implant patients with multi-electrode stimulation. Hear Res 1988;36:153-62.

21. Harkins SW. Effects of age and interstimulus interval on brainstem auditory evoked potentials. Int J Neurosci 1981; 15:107-18.

22. Stockard JJ. Nonpathologic factors influencing brainstem evoked auditory potentials. Am J Electroencephalogr 1978; 18:177-209.

23. Kileny PR, Zwolan TA, Zimmerman-Phillips S, Telian SA. Electrically evoked auditory brain-stem responses in paediatric patients with cochlear implants. Arch Otolaryngol Head Neck Surg 1994;120:1083-90.

24. Hodges AV, Ruth RA, Lambert PR, Balkany TJ. Electric auditory brain-stem responses in Nucleus multichannel cochlear implant users. Arch Otolaryngol Head Neck 1994;120:1093-9.

25. Shallop JK, Van Dyke L, Goin DW, Mischke RE. Prediction of behavioral thresholds and comfort values for Nucleus 22- channel implant patients from electrical auditory brainstem response test results. Ann Otol Rhinol Laryngol 1991;100:896-8.

26. Abbas PJ, Brown CJ. Electrically evoked auditory brainstem responses: growth of response with current level. Hear Res 1991;51:123-137.

27. Almqvist B, Harris S, Jôsson KE. The stimulogram. In Hochmair-Desoyer IJ, Hochmair ES, eds. Advance in cochlear implants. Wien: Manz; 1993:33-6.

28. Shallop JK. Objective electrophysiological measures from cochlear implant patients. In Hochmair-Desoyer IJ, Hochmair ES, eds. Advances in Cochlear Implants. Wien: Manz; 1993:21-5.

Comme le montre la figure 101, une étude plus extensive portant sur 22 sujets au lieu de 9 retrouve cette corrélation.

5 10 15 20 26 30 35

EABR Threahold (608
·10

Figure 101 : Relation entre le seuil de détection subjectif à 300 Hz et le seuil d'apparition des PEAEP à
60 Hz pour une population de 22 sujets implantés cochléaires Digisonic.

Les études utilisant d'autres types d'implant cochléaire qui ont voulu comparer les seuils de perception (à la fréquence de stimulation du réglage) aux seuils d'apparition des PEAEP (à une fréquence plus basse) n'ont pas trouvé de fortes corrélations (Brown et al 1994, Abbas et Brown, 1991 ; Mason et al, 1994, Brown et al, 1999). Le fait que le seuil de perception à 300 Hz soit très fortement corrélé au seuil d'apparition des PEAEP (R²=0.96) sur l'implant cochléaire Digisonic est intéressant. Il est possible d'utiliser les PEAEP pour le réglage de l'implant cochléaire. Ils permettent non seulement de voir si la stimulation électrique sur une électrode provoque une réponse du système auditif mais aussi de mesurer objectivement l'intensité minimale de stimulation pour laquelle le patient aura une sensation auditive (ce qui est très difficile à obtenir chez les enfants et adultes mal conditionnés). De plus le seuil d'apparition des PEAEP est très proche du seuil de perception à 300 Hz. Cela peut s'expliquer pour plusieurs raisons. Le seuil de perception sur l'implant cochléaire Digisonic est très stable pour les fréquences comprises entre 50 et 1000 Hz (cf partie psychophysique). La technique de recueil et de traitement numérique des PEAEP permet d'obtenir des PEAEP dans des conditions extrêmes (bruit physiologique et instrumental fortement réduit). La durée du pulse est utilisée pour faire varier l'intensité de stimulation ; au seuil la synchronisation est très importante car la durée du pulse est de l'ordre de 5 à 30 ps.

PEAEP et sonie

Bien que les seuils d'inconforts diffèrent en fonction de la fréquence de stimulation, nous avons étudier les relations qu'il pourrait exister entre la fonction de sonie à 300 Hz et les caractéristiques entrée-sortie des PEAEP à 75 Hz.