Hearing assessment of children is a major audiological concern, so as to early detect and manage hearing impairment, thus avoiding delayed speech development and its social, emotional, cognitive and academic hazards . Substantial progress has been made in the technology used and in the implementation of newborn hearing screening (NHS) programs worldwide . Evoked-potential audiometry (evoked with either click or tone burst stimuli) is used to measure hearing thresholds in young children and other individuals who cannot complete behavioral audiometric tests. Click Auditory Brainstem Response (ABR) thresholds correlate with behavioral thresholds at (2 - 4) kHz range (but cannot be used to reliably estimate low-frequency hearing thresholds) . The tone burst ABR (Tb-ABR), therefore, is used to evaluate frequency specific hearing threshold. More recently, Narrow Band Chirp Auditory Brainstem Response (NB-CE chirp ABR) has been available for providing hearing threshold information as a function of frequency. The chirp concept was first used in auditory electrophysiology by Shore and Nuttal . It has been studied intensively to be used in the auditory field by Elberling et al. . The CE-Chirp is a family of stimuli designed to compensate for the cochlear travel delay and provide enhanced neural synchronicity . This was done through equations based on Brainstem Auditory Evoked Potentials (BAEP) latencies recorded in humans . The broadband CE-Chirp is decomposed into four filtered narrow band chirps (NB CE-Chirp) with center frequencies 500, 1000, 2000 and 4000 Hz . This study was performed to correlate between (NB-CE chirp ABR) as a frequency specific method for hearing threshold detection in children and (Tb-ABR) in reference to behavioural hearing threshold audiometry.
2. Material and Methods
The study was carried out at audiology unit, Menoufia University hospital, from Oct. 2015 to Feb. 2017 after obtaining ethics committee approval in 2015 and informed written consent from all subjects. One hundred subjects aged 8 - 12 years of both sex with normal middle ear function were included in the study. Subjects were classified into four groups according to behavioral methods (play-conventional audiometry) as group I includes 40 subjects diagnosed as normal hearing sensitivity. Group II: included 48 subjects diagnosed as mild to moderate SNHL. Group III included 56 subjects diagnosed as moderate to moderately severe SNHL. Group IV: included 56 subjects diagnosed as severe to profound SNHL. Exclusion criteria include any neurological, mental abnormality or any other ENT problem.
All subjects in the four groups underwent pure tone audiometry using Madsen Orbiter 922 in a sound treated room amplisilence and ABR testing using (NB-CE chirp and tone burst) stimuli using interacoustic, Eclipse 25 for hearing threshold evaluation. Both narrow band Chirp stimuli and tone bursts stimuli at frequencies 500, 1000, 2000 and 4000 Hz were used with intensity 90 dBnHL down to the threshold with alternating polarity at a repetition rate 19.1 for both NB-CE-Chirp stimuli and TB stimuli. The stimuli were presented to each ear via ER3A insert phone.
2.1. Statistical Analysis
Statistical analysis was performed using SPSS version 20. One sample t-test, chi-square test and ANOVA test with post Hoc test for detection of the statistical significance between different parameters of the ABR results. P-value was considered significant when ≤ 0.05.
The age of patients ranged from 8 to 12 years. There were no statistically significant differences among the groups as regards age or sex (Table 1). Results of the study will be presented as follows; detectability of wave V till hearing threshold and comparison of wave V latency and amplitude between TB-ABR and NB-Chirp stimuli of all tested groups (Table 2 and Table 3) (Figures 1-4).
1) As regard threshold differences:
In NB CE-chirp ABR and Tb-ABR hearing thresholds were significantly overestimating behavioral thresholds. The differences were higher at 500 Hz and lower at 4000 Hz. In NB CE-chirp ABR, the difference at 500 Hz was 23.72 dB and 5.64 dB at 4000 Hz. In Tb-ABR, hearing thresholds differ by about 18.2 dB at 500 and 2.7 dB at 4000 Hz. These differences in the hearing threshold decreased with increased degree of hearing loss (Table 3) (Figures 1-4). These results agree with Federico et al.  and Almeida et al. . This may be explained by the difference in temporal integration between normal listeners and those with hearing loss. Temporal integration refers to the dependence of behavioral threshold on stimulus duration, as short-duration sounds require higher levels for detection, compared to sounds of longer duration . There were significant differences of hearing threshold levels between chirp ABR and Tb ABR at 500, 4000 Hz. No significance was shown at 1000, 2000 Hz in all groups. Rodrigues et al. have compared ABR responses evoked by NB-CE chirp and tone burst in normal hearing infants. They stated that data from hearing impaired infants can
Table 1. Age and gender distribution.
Table 2. Amplitude and latency correlation between chirp ABR and Tb ABR.
Table 3. Post HOC test between PTA, chirp ABR, Tb-ABR among 4 groups.
(p1): Comparison between PTA and CHIRP ABR. (p2): Comparison between PTA and TB ABR. (p3): Comparison between CHIRP ABR and TB ABR.
contribute to better understanding of narrow band CE-chirp stimuli in the cochlea .
2) As regards amplitude and latency difference between Tb ABR and chirp ABR:
The ABR wave responses in NB chirp ABR showed statistically higher amplitudes than Tb ABR in all frequencies (0.5, 1, 2 and 4 kHz) (Table 2). This agreed with Don et al.  and Mùhler et al.  who found that the chirp-stimulus evoked ABR showed significantly larger amplitudes than that of click-ABR. The ABR wave latencies were significantly shorter in CE-Chirp octave bands ABR than tone bursts ABR at all four frequencies. This agreed with Musiek et al.  who demonstrated wave V latencies to 60 dB nHL CE-Chirp octave band stimuli in normal-hearing young adults. These latency differences have been attributed to the input compensation technique applied to chirp stimuli in some AEP software.
3) As regards time needed to complete each test:
NB chirp ABR has less test time than Tb ABR. Also, behavioral PTA has smaller time than Chirp ABR and Tb ABR (Table 2). This agreed with Dau et al. . As chirp stimuli have short test time by aligning the arrival time of each frequency component in the stimulus to its place of maximum excitation along the basilar membrane. Such compensation will make the stimulus more efficient by achieving higher temporal synchronization between the evoked activities from the different neural elements that contribute to the formation of not only
Figure 1. Threshold difference between chirp ABR and Tb-ABR and PTA in GI.
Figure 2. Threshold difference between chirp ABR and Tb-ABR and PTA in GII.
Figure 3. Threshold difference between chirp ABR and Tb-ABR and PTA in GIII.
Figure 4. Threshold difference between chirp ABR and Tb-ABR and PTA in GIV.
ACAP but also the ABR and ASSR.
The use of a chirp-ABR testing ensures higher sensitivity and accuracy than that of Tb-ABR for measuring frequency-specific thresholds in young children. Chirp ABR consumes less time than tone burst ABR. CE-chirp ABR has larger responses than Tb-ABR responses in normal and hearing impaired children at all presentation levels. Regarding pure tone threshold prediction values of Tb-ABR stimuli, the best prediction was at 4000 Hz in normal hearing and hearing impaired children respectively, while the poorest prediction was at 500 Hz in all groups.
The authors like to thank members of audiology unit, Menoufia university hospitals for research support.
All procedures performed in the study were in accordance with the ethical standards of the institutional research committee.