The use of Lombard effect in post-laryngectomy speech rehabilitation

Abstract

This study investigated the Lombard effect, which refers to the raise in vocal intensity under noise, in laryngeal (NL) and three types of alaryngeal speakers: esophageal (SE), tracheoesophageal (TE), and electrolaryngeal (EL). A total of 52 native Cantonese laryngeal and alaryngeal speakers were recruited for the study. Speakers performed an interactive speech task involving reading a Cantonese passage under two background conditions: quiet and with 100 dB white noise. Significant increase in intensity level was found in all speaker types under noise. The intelligibility associated with all speech samples was evaluated by 15 naïve Cantonese speakers. Results showed that the overall intelligibility under noise was improved for TE and NL speakers, while no significant difference were found for SE and EL speakers across two background conditions. The relationship between the changes in jitter and shimmer under noise with intelligibility rating was revealed. Clinical implications for the use of Lombard effect in post-laryngectomy speech rehabilitation were discussed.