Journal article · Research (general)← The news desk

✦ The Dispatch

The band importance function for the Dutch digits-in-noise test: A computational derivation using human and automatic speech recognition data, including extended high frequencies

A dispatch from PubMed — filed

The band importance function (BIF) quantifies the contribution of each frequency band to the intelligibility of speech and is a key element of the speech intelligibility index. The BIF depends on the type of speech material and speaker characteristics. Several BIFs exist, but they do not cover all test materials used in audiology....

Continue reading at PubMed

✦ The floor

Discussion

Signed responses from readers of the wire.

Clinical Takeaway

This computational work refines the acoustic basis of the Dutch digits-in-noise test and suggests extended high frequencies may contribute to speech intelligibility; clinicians using this test should note these findings but no immediate protocol change is required pending clinical validation.

Why It Matters

Better understanding of which frequencies drive digits-in-noise test performance can improve test design, hearing aid fitting targets, and the interpretation of speech-in-noise scores for Dutch-speaking patients.

Key Points
  1. 01Band importance functions (BIFs) for the Dutch digits-in-noise (DIN) test were derived computationally for the first time.
  2. 02Both human listeners and automatic speech recognition (ASR) systems were used for cross-validation.
  3. 03Extended high frequencies (above 8 kHz) were found to contribute to digit recognition in noise.
  4. 04Findings could inform hearing aid frequency response targets and audiometric test design.
  5. 05Results are specific to Dutch digit stimuli and may not generalize to other languages or speech materials.
Claims & Evidence

Extended high frequencies contribute to speech intelligibility in the Dutch digits-in-noise test.

studypartially supported

Automatic speech recognition data can be used to computationally derive band importance functions comparable to human listener data.

studypartially supported
Research metadata
PMID
42673779
DOI
10.1016/j.heares.2026.109785.
Journal
Hearing Research
Publication type
research_article
Evidence level
2b
Population
Human listeners and automatic speech recognition systems tested on Dutch digit stimuli
Intervention
Computational derivation of band importance function using human and ASR speech recognition data including extended high frequencies
Comparator
Human speech recognition data vs. automatic speech recognition data

Primary outcomes

Band importance function values per frequency band; Contribution of extended high frequencies to digit recognition in noise

Related stories