The simplified French Matrix. A tool for evaluation of speech intelligibility in noise

Abstract

Introduction: In audiology, the usual tests of speech perception in silence are non-predictive of intelligibility in noise. The French Matrix was developed to assess intelligibility in noise in adults with normal cognitive capacity. A simplified adaptive version, FRA-SIMAT, was derived for use with children and elderly persons with diminished memory span. The aim of the present study was to apply this adaptive procedure to determine signal-to-noise ratio (SNR) according to percentage intelligibility.

Methods: Twenty normal-hearing adults and 60 normal-hearing children aged 5-6 years (G1), 7-8 years (G2) and 9-10 years (G3) were included. FRA-SIMAT uses groups of 3 words, to limit memory demand.

Results: In adults, SNR for 50% (SNR-50) and 80% intelligibility (SNR-80) was respectively -7.1±1.4dB and -3.7±1.6dB. In children, SNR-50 was -4.2±1.3dB in G1, -4.6±1.2dB in G2, and -5.5±1.5dB in G3.

Conclusion: The FRA-SIMAT test of speech perception in noise was validated in adults and children of different ages, and can be implemented as a diagnostic and rehabilitation tool in clinical practice.

Keywords: Adaptive test of speech intelligibility in noise; SNR.