Auditory Brainstem Response
Design, run, and score ABR testing — prep electrodes, acquire waveforms across intensities, mark waves I/III/V, estimate Wave V threshold, and interpret latencies.
The Auditory Brainstem Response element simulates clinical ABR testing. The learner preps the electrodes, acquires evoked waveforms at a range of intensities, marks Waves I, III, and V, estimates the Wave V threshold, measures interpeak latencies, and interprets the pattern (normal, conductive, cochlear, retrocochlear, mixed, or no response). It supports click, tone-burst (500 / 1000 / 2000 / 4000 Hz), and CE-Chirp stimuli, per ear. In the designer it's listed as:
Electrode prep, then mark waves I/III/V and interpret latencies — click / tone / chirp, per ear
Like every Theta element, it follows the lifecycle Design → Settings → Testing → Scoring → Report.
Design
You design the ABR per ear, choosing a pathology and the protocol the learner may use to test it.
Pathology per ear
Pick a category for each ear; the relevant magnitude fields appear:
| Pathology | Fields | Effect |
|---|---|---|
| Normal | — | Baseline latencies and morphology. |
| Conductive | Air-bone gap (dB) | Delays absolute latencies; interpeaks stay normal. |
| Sensorineural (cochlear) | Cochlear loss (dB) | Raises Wave V threshold, degrades morphology, weakens early waves. |
| Retrocochlear | Retrocochlear delay (ms) | Delays Wave V (and partly III), prolonging interpeaks. |
| Mixed | Air-bone gap + cochlear loss | Both effects together. |
In template mode each ear gets its own randomization ranges, set right under the threshold table:
- Severity offset — one random dB shift applied to every threshold, so the audiometric shape and the loss type survive while the case severity varies.
- Air-bone gap offset — shifts the conductive component. It only applies where you designed a gap, so a case built without one never grows one.
- Retrocochlear delay — the range the Wave V delay is re-rolled within (retrocochlear ears only).
In fixed mode the exact values are used every time. If the element is linked to an audiogram, that audiogram's template drives the variation instead and these ranges are ignored.
Optional audiogram link
You can link the ABR to an existing Audiogram element in the lesson. When linked, thresholds and air-bone gaps are derived from that audiogram — the 2–4 kHz average for click / CE-Chirp, or the specific frequency for tone bursts. The cochlear and retrocochlear components always come from the ABR fields above.
Protocol
The protocol defines what the learner can acquire:
- Ears — Right and/or Left
- Stimuli — Click, Tone burst (+ which frequencies), CE-Chirp
- Polarities — Rarefaction, Condensation, Alternating
- Rate (default 11.1 /s) and Sweeps / averages (default 2000)
- Intensity range — Min (default 20), Max (default 80), Step (default 10) dB nHL
The step also drives the learner's level control, so a 10 dB step means the console moves in 10 dB. At least one option always stays selected in each list — a protocol with no ears or no stimuli would leave a console nobody can drive.
A live preview shows the answer-key waveforms and expected interpretation for the selected ear and stimulus, at a rate, polarity and sweep count you choose — so you can see the rate and polarity effects before a learner ever runs the case.
Simulated factors (rate, retrocochlear, sweeps, polarity) switch those effects off in the waveform model to simplify a teaching case. The console still offers the control; the trace simply stops responding to it.
Settings
The ABR Settings carry the normative model and tolerances that are frozen into each submission for reproducible scoring.
| Setting | Default | Notes |
|---|---|---|
| Normative Wave V latency | 5.6 ms | At 80 dB nHL, click, normal ear. Waves I/III/V ≈ 1.5 / 3.7 / 5.6 ms. |
| Latency–intensity slope | 0.03 ms/dB | Equal shift per wave, so interpeaks stay stable. |
| Interpeak normal maxima | I–III 2.5 · III–V 2.4 · I–V 4.5 ms | Above these, the answer key calls the interpeaks prolonged. |
| Absolute latency allowance | 0.4 ms | How far past the normal-ear value Wave V may sit before the answer key calls the absolute latencies prolonged. |
| Impedance target / balance | 5 kΩ / 3 kΩ | Each electrode must reach the target; the spread must stay within balance. |
The expected interpretation is read off the same generated waveform the learner marks, so changing these norms changes the answer key with them.
Marking tolerances are not here — they live on the Scoring tab, which is what the scorer reads.
Testing
A submission has two phases.
1. Electrode prep. The learner scrubs each of four electrode sites (Cz, A1, A2, Fpz) to lower its impedance. Scrubbing drives the reading down toward a floor; the learner sees each site's current value but not the target — whether the montage is acceptable is checked silently at scoring time.
2. Testing. In the console the learner selects ear, stimulus, level, rate, polarity, and sweeps, then Acquire Trace. Acquired runs stack high-to-low by intensity. The learner:
- Clicks a trace to place the selected wave (I / III / V, plus II/IV for reference). Morphology degrades near threshold, at fast rates, with few sweeps, and with retrocochlear delay.
- Reads the latency–intensity function (Wave V latency vs level) to estimate the Wave V threshold.
- Records the interpretation: absolute latencies (normal / prolonged), interpeak latencies (normal / prolonged), and a conclusion (normal, conductive, cochlear, retrocochlear, mixed, or no response).
Tapping a trace both places the marker and selects that trace. Clear removes the wave from the selected trace only; Clear — all removes it from every trace on screen.
If the case keeps the prep step, recordings made before the electrodes are checked are noisy and poorly formed — the same as testing on unprepped electrodes. Cases set to Assume normal impedances skip prep entirely and always record cleanly.
Scoring
Scoring is a tri-state per field, per ear — Scored, Report-only, or Off — with per-ear tolerances (peak ± 0.3 ms, interpeak ± 0.4 ms, threshold ± 10 dB by default).
Peaks are graded on the highest-level run of each stimulus the learner tested — the condition waves are read from clinically. Every wave genuinely present there must be marked: leaving one out is scored wrong, not skipped. Marks on the other runs of that stimulus are shown on the report but never counted, so collecting extra levels can only help. Two learners who tested the same stimuli are therefore graded out of the same number of points.
| Field | What's compared |
|---|---|
| Peak marks (I / III / V) | Marked latency vs regenerated ground truth, within peak tolerance. |
| Interpeaks (I–III, III–V, I–V) | Computed from the learner's marks vs the truth, within interpeak tolerance. |
| Wave V threshold | Learner's estimate vs the true threshold, within threshold tolerance. |
| Absolute / interpeak judgments | Exact match of normal / prolonged. |
| Conclusion | Exact match of the pathology category. |
| Electrode impedances | All sites ≤ target and spread ≤ balance (element-wide, checked silently). |
Rows set to Off are excluded; Report-only rows appear on the report but don't count; Scored rows are each correct/incorrect and sum to the element's points.
Report
On the PDF, each ear shows:
- A waveform stack per stimulus tested (up to three, highest-intensity traces first), with the learner's Wave I / III / V marks overlaid. The traces are redrawn with the exact acquisition settings the learner used, so the report shows the recording they marked — noise, morphology and all. Anything the page can't fit is named in the caption rather than dropped silently.
- A latency table — one row per acquired run with the I / III / V latencies and the computed interpeaks.
- An interpretation summary — the learner's absolute-latency, interpeak, and conclusion calls next to the expected values.