Development of a software for evaluation of the spatial processing of sentences in noise.

Introduction: In order to understand speech in a noisy environment, it is necessary to focus on the person we want to hear and simultaneously suppress competing sounds from different places. This process is called Spatial Processing. However, tests to assess this ability are only available in English. Objectives: To develop a Spatial Processing with Sentences in Noise test (PROSER) in Portuguese and to establish normality criteria. Methods: The research was approved by the Institution's ethical committee under No. 3,462,572. It was developed in four stages: creation of a software to apply PROSER; selection, recording and editing of 188 sentences and two children's stories; testing all sentences to equalize them for equal intelligibility; application of the test in 48 children (7 to 10 years) with typical development to determine the criteria of normality. Results: The PROSER test evaluates the understanding of sentences (presented in the 0° azimuth condition) in four variations of competitive speech presentation (stories): Same Voice (0°), Same Voice (±90°), Different Voices (0°) and Different Voices (±90°). The test evaluates the performance of each individual in regards to its Speech Recognition Threshold in the four different conditions. The values generated in the calculation of the advantages represents the benefit in decibels obtained when cues of different voices, spatial or both will be incorporated in the competitive message, compared with the result of the Same Voice 0° condition. Conclusion: The software and speech material in Portuguese are finished and normality criteria for Spatial Processing in Brazilian children have been established. [ABSTRACT FROM AUTHOR]