Visualise the Auditory Brainstem Response — the brainstem-generated scalp potential that follows an acoustic click within 10 ms. Adjust stimulus level, click rate, sweep count and patient condition to see Wave I-V latencies, Inter-Peak Intervals, sqrt(N) SNR improvement and the estimated ABR threshold update in real time.
Parameters
Stimulus level
dB nHL
Click intensity in dB normal Hearing Level
Click repetition rate
Hz
Higher rates can delay Wave V through brainstem adaptation
Number of sweeps N
Coherent averages — SNR improves as sqrt(N)
Stimulus type
Click is broadband; tone-bursts estimate frequency-specific thresholds
Patient condition
Adjusts latency correction and estimated threshold
Ipsilateral ear
Side of the recorded ipsilateral channel
Results
—
Wave I latency (ms)
—
Wave V latency (ms)
—
IPI I-V (ms)
—
SNR gain (dB)
—
Estimated threshold (dB nHL)
—
Test time (s)
—
ABR waveform — Wave I-V markers and averaging progress
Brainstem potential recorded within 10 ms of the click (Cz vs mastoid). The lower bar shows the averaging progress; the noise envelope shrinks as N grows.
L is the stimulus level (dB nHL) and N the number of averaged sweeps. Latency shortens by 1.0 ms per +10 dB, and the SNR improves as sqrt(N) with coherent averaging.
The minimum level at which Wave V disappears estimates the audiometric threshold. Typical shifts are +35 dB for conductive and +55 dB for sensorineural loss.
Auditory Brainstem Response (ABR) — hearing screening and latency diagnosis
🙋
I heard that newborn babies get a hearing test right after birth. But babies cannot tell you "I can hear" yet — how is the test actually done?
🎓
Good question. Because newborns cannot respond, we directly measure whether the auditory signal reaches the brain using electrodes stuck to the scalp. That is the ABR — Auditory Brainstem Response. We deliver a sharp "click" at about 20 per second next to the ear and record the few-hundred-nanovolt potential that appears within 10 ms, averaging 1000-2000 sweeps. Out comes a clean train of five peaks called Waves I through V. Automated newborn screening (AABR) simply asks whether Wave V is reliably present.
🙋
Five peaks — that's a lot! What does each one stand for?
🎓
Each peak comes from a different generator along the ascending pathway. Wave I is the distal cochlear nerve, then Wave II at the proximal nerve, Wave III at the cochlear nucleus once the signal enters the brainstem, Wave IV at the superior olivary complex, and Wave V near the inferior colliculus. The peaks roughly line up at 1, 2, 3, 4 and 5 ms in a healthy adult. Try lowering the stimulus level to 60 dB on the left slider — every wave should shift right by 0.05 × (80 − 60) = 1.0 ms. That shift is exactly the Latency-Intensity Function of ABR.
🙋
OK! And what happens when I switch the patient condition to "retrocochlear lesion"?
🎓
A retrocochlear lesion — say a vestibular schwannoma (acoustic neuroma) — slows conduction inside the brainstem. So Wave I still looks normal but Waves III and V arrive late, and the Inter-Peak Interval I-V (IPI I-V) stretches. A healthy adult has about 4.0 ms; above 4.6 ms is a red flag. Look at the "IPI by condition" chart below: a pure sensorineural loss barely changes the IPI, but the retrocochlear case clearly pushes IPI I-V to the right. Even with MRI everywhere, ABR remains a cheap and radiation-free way to screen for retrocochlear lesions.
🙋
When I crank up the sweep count, the "SNR gain" shoots up. Why does averaging kill the noise?
🎓
Because the spontaneous EEG and EMG are random with respect to the stimulus, while the ABR is time-locked to it. Average N sweeps and the signal adds linearly while the noise grows only as sqrt(N), so the SNR improves as sqrt(N). For N=1500 the gain is 10·log10(1500) ≈ 31.8 dB — exactly enough to pull the ABR out of the background EEG. That is why 1000-2000 sweeps is standard. More sweeps would help further, but the test gets longer and a fussy baby with strong EMG can defeat the gain. Hitting "enough SNR in a short time" is the art of ABR testing.
🙋
What is the difference between a "click" and a "tone-burst"? Which one do I use if I want a frequency-specific threshold?
🎓
A click is a 0.1 ms sharp pulse with broadband energy from 1 to 4 kHz, so it can only tell you whether the mid-to-high frequencies roughly work. Tone-bursts (500 Hz, 1 kHz, 2 kHz, 4 kHz) are short tones centred on a chosen frequency, so they give frequency-specific thresholds — essential for fitting a hearing aid. Standard practice is to screen with a click first, then for Refer babies run tone-burst ABR or ASSR (Auditory Steady-State Response) to draw a full audiogram before fitting amplification.
Frequently Asked Questions
Each ABR wave reflects an electrical generator along the ascending auditory pathway. Wave I arises from the distal cochlear nerve, Wave II from the proximal cochlear nerve, Wave III from the cochlear nucleus, Wave IV from the superior olivary complex, and Wave V from the lateral lemniscus / inferior colliculus region. Clinically, the most reproducible Waves I and V are used to evaluate latency, amplitude and Inter-Peak Interval. Typical adult latencies are I=1.5, III=3.5 and V=5.5 ms for an 80 dB nHL click.
The IPI reflects brainstem conduction time and is largely insensitive to peripheral hearing loss, which makes it useful for detecting retrocochlear lesions. In adults, IPI I-V is typically around 4.0 ms, with IPI I-III and III-V around 2.0 ms each. An IPI I-V above 4.4-4.6 ms or a left-right difference above 0.2-0.3 ms suggests a retrocochlear lesion such as vestibular schwannoma (acoustic neuroma). In this tool, selecting the 'retro_cochlear' condition slows brainstem conduction and stretches the IPI accordingly.
ABR scalp potentials are only a few hundred nanovolts, buried in microvolt-level background EEG and EMG noise. Stimulus-locked averaging across N sweeps adds the time-locked signal coherently while the random noise cancels in proportion to sqrt(N), so the SNR improves as sqrt(N). For N=1500 the gain is 10·log10(1500) ≈ 31.8 dB. That is why 1000-2000 sweeps is the clinical standard.
Automated ABR (AABR) uses a fixed click level around 35 dB nHL and only reports a Pass / Refer decision based on automatic Wave V detection. A diagnostic ABR sweeps the stimulus level from 20 to 100 dB nHL, evaluates the Wave V latency-intensity function, IPI and frequency-specific thresholds with tone-bursts. Diagnostic ABR is what confirms the presence, site (conductive / sensorineural / retrocochlear) and threshold of hearing loss after a Refer at AABR screening.
Real-World Applications
Newborn hearing screening (AABR): In many countries more than 90% of newborns are screened with AABR before discharge. A 35 dB nHL click is delivered and the device automatically decides Pass or Refer based on the reproducibility of Wave V. Early detection and intervention — hearing-aid fitting before 6 months of age — has been shown to dramatically improve language outcomes in bilaterally deaf infants. Try lowering the stimulus level to 35 dB in this tool to feel how borderline Wave V becomes at screening intensities.
Diagnosis of vestibular schwannoma: Before MRI became routine, ABR was the first-line screen for retrocochlear pathology. It is still used today when MRI is contraindicated (metallic implants) or to follow up unilateral sensorineural hearing loss. Positive findings include IPI I-V > 4.6 ms, left-right difference > 0.3 ms and a reduced Wave V amplitude. Choosing 'retro_cochlear' here stretches IPI I-V to about 4.8 ms and pushes the verdict to "ng".
Intra-operative auditory monitoring (IOM): During acoustic neuroma resection or cerebellopontine angle surgery, the auditory nerve and brainstem are monitored continuously to preserve hearing. A Wave V latency increase of more than 1 ms or an amplitude drop of more than 50% triggers an alert to the surgeon. Producing one averaged trace every 30-60 seconds, the balance between speed and reliability is what defines a good IOM team.
Coma assessment and brain-death adjunct testing: ABR is used as an adjunct to evaluate brainstem function in comatose patients and as part of brain-death protocols. In brain death the brainstem-generated Waves II-V vanish and only the peripheral Wave I remains. In partial brainstem injury, selective loss of Wave V may help localise the lesion.
Common Misconceptions and Pitfalls
The biggest pitfall is equating "AABR Pass" with "normal hearing". AABR usually decides Pass / Refer at a fixed 35 dB nHL, so it cannot detect mild hearing loss (20-35 dB HL) or a low-frequency-only loss. Because the click stimulus is dominated by 1-4 kHz energy, a baby with a low-frequency loss below 500 Hz can still pass. If a child shows speech-language delay or poor responsiveness after a Pass, always follow up with tone-burst ABR, ASSR or behavioural audiometry — "passed the screen, no need to worry" is wrong.
Next, assuming that higher click rates always shorten the test. At 90 Hz, 1500 sweeps finish in about 17 seconds, but brainstem adaptation slows Wave V by 0.5-1.0 ms and the normal latency tables no longer apply. The clinical standard is 11-21 Hz; "rate-study" protocols deliberately raise the rate to stress the brainstem, but only against rate-specific normative data. This tool simplifies the rate effect, so always cross-check against your lab's normative tables when interpreting real recordings.
Finally, the belief that "the lower the electrode impedance the better". High impedance (>5 kΩ) does raise noise, but aggressively scrubbing the skin to push it below 1 kΩ can hurt the baby, and large impedance imbalance between the two recording electrodes ruins common-mode rejection and actually increases noise. The ideal is "all three electrodes balanced between 3-5 kΩ", especially in neonates where gentle skin prep is essential. The sqrt(N) SNR figure here is an ideal-world value; in practice electrode condition, infant movement and ambient electrical interference all push you off the sqrt(N) curve.
How to Use
Set stimulus level between 60-90 dB nHL using the stimulusLevelDB slider to control click intensity delivered via insert earphones
Adjust click rate from 10-90 Hz via clickRateHz control; higher rates (70+ Hz) improve SNR but reduce wave morphology clarity
Configure numSweeps from 1000-4000 repetitions; 2000 sweeps typical for diagnostic ABR yields adequate signal-to-noise ratio at 90 dB nHL
Run simulator to generate latency measurements (Wave I, Wave V), interpeak interval I-V, SNR gain in dB, and estimated auditory threshold
Worked Example
Clinical ABR protocol at 80 dB nHL: stimulus level 80 dB, click rate 21.1 Hz (standard), 2048 sweeps. Expected outputs: Wave I latency 1.6 ms, Wave V latency 5.8 ms, IPI I-V 4.2 ms (normal <4.4 ms), SNR gain approximately 18-22 dB after 2048 averages. Test duration approximately 97 seconds. Threshold estimation via descending level protocol (10 dB steps from 90 dB nHL) identifies auditory brainstem function; absent Wave V above 90 dB nHL indicates retrocochlear pathology or profound hearing loss.
Practical Notes
Higher click rates (80 Hz) reduce test time but increase baseline noise floor—use 21.1 Hz for diagnostic sensitivity in suspected acoustic neuromas where interaural latency difference exceeds 0.3 ms
Infant ABR requires 50+ dB nHL stimulus intensity and 4000+ sweeps due to immature neural generators; adult protocols use 70-90 dB nHL
SNR gain improves by 3 dB per doubling of sweeps; 4000 sweeps yields ~6 dB advantage over 1000 sweeps but extends session from 47 to 190 seconds
Wave V morphology degrades at stimulus levels below 60 dB nHL; always confirm presence at suprathreshold levels before threshold estimation