HK Audiology Group
Tool governance

Tool Version Log

A public baseline for NeuraHear assessment tools, release notes, data-boundary statements, and future study-facing changes.

Why this matters

Research tools change over time. Version notes help future users know which calibration, stimuli, scoring, export, and data-handling assumptions applied.

Back to tools

Document the current baseline

Each public research tool should show its stage, status, data handling, outputs, and version note before use.

Separate interface updates from study data

Static site releases should not overwrite server-backed records such as iDIN participant results.

Record meaningful behavior changes

Calibration, stimuli, scoring, export format, and backend changes should be logged before a tool is used for new data collection.

Current baseline

Public assessment-tool register

LiveSpeech-in-noise assessment

integrated Digit-in-Noise Test

A browser-based screening workflow with calibration, device checks, adaptive SNR testing, and server-backed data capture.

Version note
Reliable result upload
Data boundary
Responses are submitted to the project backend for study use.
LiveStimulus optimization

Digit Optimization Test

A fixed-SNR single-digit workflow for estimating PI functions and generating correction levels for sequence synthesis.

Version note
Noise calibration enabled
Data boundary
Runs in the browser; individual and pooled results are exported by the researcher.
In validationEveryday functioning PROM

HFEQ-Mandarin Research Preview

An ICF-based questionnaire workflow for Mandarin-speaking adults, covering hearing, communication, participation, support, personal resources, and health.

Version note
Manuscript-linked preview
Data boundary
Browser-only completion and export for research preview use.
In developmentAI in hearing care

AI-Guided Hearing-Care Workflows

Planned tools for AI-supported questionnaire guidance, triage, interpretation, and care-pathway support, developed with human oversight.

Version note
Concept stage
Data boundary
No public tool yet; future tools will separate research data, generated outputs, and review records.

Release notes

Recent platform and tool-facing changes

Released2026-06-11integrated Digit-in-Noise Test

iDIN reliable result upload added

iDIN 2026.06 upload queue

The iDIN results page now protects completed test records from being lost when server upload is interrupted.

  • Added a browser-side pending upload record before the first server save attempt.
  • Added clear Saved, Saving, Pending, and Failed states on the results page with a Retry upload action.
  • Stored a stable uploadClientId with each completed session for audit and duplicate protection.
  • Added backend idempotency so repeated uploads with the same uploadClientId or sessionId return the existing record instead of appending another JSONL row.
  • Added upload metadata to researcher dashboard detail view and summary CSV export.

Data boundary: Result upload reliability and metadata changed only; audio files, fixed-noise-gain application, SNR logic, adaptive scoring, randomization behavior, and existing server-side result records are unchanged.

View related page
Released2026-06-11integrated Digit-in-Noise Test

iDIN test interaction state safeguards added

iDIN 2026.06 test-state UX

The iDIN test page now uses clearer playback and response states so participants cannot enter responses before audio playback is complete.

  • Added a compact state badge for Ready, Preparing, Listening, Enter response, Saving, and playback error states.
  • Blocked physical-keyboard digit entry before playback has finished, matching the disabled on-screen keypad behavior.
  • Kept the response keypad visually disabled in ready/error states while preserving Play retry behavior.
  • Changed GitHub Actions so pushes run lint/build checks only; production deployment now runs only when the workflow is manually dispatched, avoiding overlap with manual safe deploys.

Data boundary: Front-end interaction safeguards and deployment trigger behavior changed only; audio files, fixed-noise-gain application, SNR logic, adaptive scoring, randomization behavior, and existing server-side result records are unchanged.

View related page
Released2026-06-11integrated Digit-in-Noise Test

iDIN calibration confirmation guard added

iDIN 2026.06 calibration-confirmation UX

The iDIN calibration flow now requires participants to play the noise and explicitly confirm audibility before starting the formal test.

  • Added a required audibility confirmation after noise playback on the calibration page.
  • Disabled test start until audio is loaded, noise has been played at least once, and the calibrated level is confirmed.
  • Reset the audibility confirmation when participants adjust the noise level after confirming.
  • Stored calibration quality-control metadata including play count, first playback timestamp, cumulative noise-listening time, and adjustment count.

Data boundary: Front-end calibration gating and metadata changed only; audio files, fixed-noise-gain application, SNR logic, adaptive scoring, randomization behavior, and existing server-side result records are unchanged.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN pre-test readiness check added

iDIN 2026.06 readiness-check UX

The iDIN setup flow now asks participants to confirm key device and environment conditions before calibration begins.

  • Added required headphone, quiet-environment, and browser-audio readiness confirmations before calibration.
  • Stored readiness confirmations and timestamp in participant setup metadata and result meta.
  • Added inline validation so participants cannot continue until readiness checks are confirmed.

Data boundary: Front-end setup metadata changed only; audio files, SNR logic, adaptive scoring, randomization behavior, and existing server-side result records are unchanged.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN calibration audio readiness added

iDIN 2026.06 calibration ready-state UX

The calibration page now makes audio loading explicit and prevents users from pressing calibration controls before noise playback is ready.

  • Added loading, ready, playing, stopped, and error states to the calibration page.
  • Disabled Play noise and Confirm until the selected language audio has finished loading.
  • Added loading progress feedback for the 10 digit files plus the noise file.
  • Resumed the browser audio context on Play noise so browser autoplay protections do not silently block calibration playback.
  • Added timeout recovery so calibration playback cannot remain indefinitely in the starting state.

Data boundary: Front-end calibration interaction changed only; audio files, SNR logic, adaptive scoring, randomization behavior, and existing server-side result records are unchanged.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN test-page layout stabilized

iDIN 2026.06 stable keypad UX

The iDIN landing and test pages now use more consistent controls, with a stabilized response keypad that stays in place during playback and answer entry.

  • Unified the iDIN landing-page test launch and researcher dashboard controls into one action group.
  • Separated changing playback/status messages from the progress row and reserved stable layout space for the test interaction.
  • Refined the keypad backspace control with a cleaner icon and clearer button state.
  • Bumped the iDIN static asset version so deployed browsers fetch the updated interface.

Data boundary: Front-end interaction and styling changed only; audio mixing, SNR logic, adaptive scoring, randomization behavior, and existing server-side result records are unchanged.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN calibration and dashboard workflow refined

iDIN 2026.06 flow and dashboard UX

The iDIN workflow now has clearer noise calibration controls, smoother in-test transition messages, and stronger researcher dashboard filtering and export tools.

  • Added calibration gain and dB readouts, direct dB entry, -1 dB/+1 dB controls, and reset while preserving the existing fixed-noise-gain algorithm.
  • Replaced normal practice/formal transition alerts with inline status messages and short automatic continuation delays.
  • Added dashboard filters, QC summary tiles, and current-list CSV export for researcher review.

Data boundary: Front-end workflow and dashboard UI changed only; audio mixing formulas, adaptive scoring, and existing server-side result records are unchanged.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN sequence-level SNR definition documented

iDIN 2026.06 SNR documentation

The iDIN implementation notes now explicitly define SNR matching as a corrected sequence-level operation, not per-digit SNR matching.

  • Documented that digit waveforms are RMS-normalized before digit-specific correction levels are applied.
  • Clarified that corrected digits are synthesized into a sequence before global SNR matching against noise.
  • Clarified that identical digits may have different realized single-digit SNRs across sequences, by design.

Data boundary: Documentation and version metadata changed only; the mixing algorithm, adaptive rule, and existing server-side results are unchanged.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN setup and dashboard wording refined

iDIN 2026.06 setup refinement

The integrated DIN setup page now requires participant IDs, and the researcher dashboard has fully English action wording.

  • Made Participant ID a required setup field with both browser-level and script-level validation.
  • Changed the dashboard result action button from Chinese wording to English.
  • Bumped the iDIN static asset version so deployed browsers pick up the updated files.

Data boundary: Static iDIN front-end files changed only; existing server-side JSONL results and backend storage are not modified by this release.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN condition order randomization added

iDIN 2026.06 condition-order release

The integrated DIN setup page now supports optional randomization of selected test conditions and exposes a researcher dashboard entry from the iDIN landing page.

  • Added a researcher dashboard link on the iDIN landing page.
  • Added an optional randomize-condition-order checkbox during participant setup.
  • Stored the original selected condition order, executed condition order, randomization flag, and generation timestamp in the result metadata.

Data boundary: Static iDIN front-end files changed only; existing server-side JSONL results and backend storage are not modified by this release.

View related page
Released2026-06-10integrated Digit-in-Noise Test

iDIN playback and audio asset baseline tightened

iDIN 2026.06 stability release

The integrated DIN workflow now has stronger playback-state handling, cache-busted static assets, clearer formal/practice export metadata, and shorter noise files for faster loading.

  • Prevented repeated Play/OK actions from creating unstable trial states.
  • Separated formal-trial metrics from practice-trial records in result summaries and exports.
  • Trimmed long language noise files to approximately 30 seconds without changing sample rate, channel count, bit depth, or digit stimuli.
  • Added versioned CSS/JS references so browsers fetch the current iDIN interface after deployment.

Data boundary: Static tool files and audio assets changed; the server-side iDIN results JSONL data directory is not modified by this release.

View related page
Released2026-05-20NeuraHear platform

Public version log established

Site release 2026.05.20

A public version and credibility layer now records the current tool baseline, release notes, contact route, and data-boundary statements.

  • Added a research contact page and connected it to navigation, footer, sitemap, and structured data.
  • Added tool study-planning cues so researchers can check versioning, calibration, export format, and backup expectations before use.
  • Established this version log as the place to record future tool-facing releases.

Data boundary: Static website content only; the iDIN server data directory and study records are not modified by this release.

View related page
Baseline2026-05-20Assessment tools

Current assessment-tool baseline captured

Baseline 2026.05

The public tool registry now makes each tool's stage, validation status, data handling, outputs, and version note visible from the Tools page.

  • iDIN remains the operational server-backed speech-in-noise workflow.
  • Digit Optimization is listed as a live stimulus-development workflow with noise calibration enabled.
  • HFEQ-Mandarin is listed as an in-validation PROM research preview.

Data boundary: The baseline is a documentation snapshot; individual tool behavior remains governed by each tool page and deployment route.

View related page
Released2026-05-20NeuraHear platform

Discovery metadata and sitemap hardened

SEO release 2026.05

The site now publishes route metadata, canonical URLs, robots rules, sitemap entries, and social-card metadata for better indexing and sharing.

  • Added site-wide metadata defaults and page-level metadata helpers.
  • Added robots and sitemap generation for primary public routes.
  • Added public social-card metadata for shared links.

Data boundary: Static discovery metadata only; no participant records or test data are touched.

View related page

Before a study release

Log behavior changes before new data collection starts.

Calibration flow, audio stimuli, scoring, export fields, server routes, and backup assumptions should be recorded here before a protocol uses a new version.

Study workflow