Intelligibility, subjective ratings and completion time scores using the FAAF test with hearing-impaired subjects and noisy reverberant environments

• Published in: British journal of audiology Vol. 35; no. 4; pp. 237 - 245
• Main Authors: Shields, P.W., Campbell, D.R.
• Format: Journal Article
• Language: English
• Published: London Taylor & Francis 01.08.2001; Whurr
• Subjects: Adult; Aged; Biological and medical sciences; Ear, auditive nerve, cochleovestibular tract, facial nerve: diseases, semeiology; FAAF test; Female; Four Alternative Auditory Features test; Hearing impaired people; Hearing Loss, Sensorineural - diagnosis; Hearing Tests; hearing-impaired subjects; Humans; Intelligibility; intelligibility testing; Investigative techniques, diagnostic techniques (general aspects); Male; Medical sciences; Middle Aged; Noise - adverse effects; Noisy environment; noisy reverberant environments; Non tumoral diseases; Otorhinolaryngology functional investigation (larynx, voice, audiometry, vestibular function, equilibration...); Otorhinolaryngology. Stomatology; Perceptual Masking - physiology; Severity of Illness Index; Speech; Speech Perception - physiology; Time Factors; Human; Performance evaluation; Acoustic reverberation; Statistical analysis; Healthy subject; Intelligibility; Auditory disorder; Background noise; Experimental protocol; Speech recognition; ENT disease; Adult; Perception hearing loss; Diagnosis; Technique; Comparative study
• ISSN: 0300-5364
• DOI: 10.1080/00305364.2001.11745242

A series of experiments have been performed with the primary aim of assessing the performance of a signal-processing algorithm for a possible future hearing aid application. As part of this work the four alternative auditory feature (FAAF) test was used to obtain a quantitative assessment of speech intelligibility and a subjective assessment of speech quality. This paper reports results of experiments using normal hearing (NH) subjects that provide partial verification of the FAAF test originators' prior work. Also reported are intelligibility score, mean opinion score (MOS) and completion time data obtained by use of the FAAF test in experiments using hearing-impaired (HI) subjects with moderate sensorineural hearing loss, experiencing a speech signal corrupted by additive speech-shaped or multi-talker babble masking noise, and either simulated or real reverberation. Over the practically important signal-to-noise ratio (SNR) range from -6 dB to +9 dB, and using data recorded in reverberant surroundings, the FAAF test allowed moderately HI subjects using amplification according to National Acoustic Laboratories (NAL) recommendations to demonstrate an improvement in speech intelligibility as the SNR increased. This was strongly correlated with a perceived improvement in speech quality and a decrease in time taken to complete the test. The data obtained provide some measures of the relationship between time to complete a FAAF test and SNR, MOS and SNR, and intelligibility score and SNR. In addition, the advisability of relying solely on the simple 'image method' of simulating an acoustic environment is questioned on the basis of data presented. These data derive from an experiment with a particular engineering aim, and while neither the experiment nor the results arc claimed to be complete enough to make strong audiological deductions, they should be of practical support to experimenters planning to use the FAAF test with NH or moderately HI subjects, complex acoustics and masking noise.