Automated Measures of Speech Intelligibility Using Brain Wave Recordings

Automated Measures of Speech Intelligibility Using Brain Wave Recordings

Hearing loss is one of the most prevalent impairments in the world. The World Health Organization estimates that as of 2020, around 466 million people worldwide are affected by disabling hearing loss, of which 34 million are children. Without appropriate diagnosis and treatment, hearing impairment can cause difficulties with communication, learning, social-emotional functioning, employment and quality of life. Early intervention is especially important for children, for whom unaddressed hearing loss has been proven to affect speech and language development, educational attainments, and social skills. Through early detection and interventions many of these impacts can be mitigated, highlighting the importance of accurate hearing diagnostics.

Evaluation of speech intelligibility is a fundamental component of hearing assessment and rehabilitation. The current gold standard in measuring speech intelligibility relies heavily on behavioural tests. While these tests are reliable and fast in healthy adults, it is challenging to assess speech intelligibility for patients who cannot communicate clearly, such as young children. For example, behavioral tests require active participation from the patient, such as raising their hand when a sound is heard, or repeating words and sentences.

Objective measures, for instance using electro-encephalography (EEG), could overcome the current challenges in hearing assessment and do not require active participation from the patient. For example, in Flanders, all newborns are screened with an objective test where a few sensors on the head detect brainwaves in response to sounds. However, objective tests only use simple sound stimuli, such as tones or clicks, which cannot be used to measure important high-level hearing outcomes, such as speech understanding. In other words, current objective tests are used to check if a person can hear a word, but they can't indicate if the person can understand the word.

An obvious missing link in audiological practice is a diagnostic test that can measure high-level hearing outcomes such as speech understanding in an objective way. For this reason, CORGEE was developed. CORGEE uses a novel method of "neural speech tracking" to objectively measure speech intelligibility using EEG. In the current study, the efficacy and the clinical validity of the CORGEE software will be evaluated in young children with a hearing impairment.

This study is set-up as a pre-market clinical investigation of a non-CE marked medical device. The main goal of the study is to measure the efficacy of the CORGEE software, a newly developed software application that will function as an added service in the auditory assessment and counselling by hearing care professionals for persons with (suspected) hearing-related problems.

We will evaluate the efficacy and the clinical validity of the CORGEE software by means of a within-subject cross-sectional study in infants, toddlers, kindergartners and primary school children. The study will be carried out in 90 children with (suspected) hearing impairment. Participants are recruited via four participating sites (one university hospital and three ambulatory revalidation centers) across a period of 12 months.

All participants will undergo an objective assessment of speech intelligibility based on electrophysiological responses to natural running speech, performed by their healthcare professional (HCP). Using the CORGEE device, neural tracking of the speech envelope will be investigated with and without hearing aids at different levels of speech intelligibility. With this clinical study, we aim to demonstrate an objective hearing aid benefit when listening to natural speech using EEG. Moreover, we aim to compare the objective results with behavioural measures of speech intelligibility obtained with standard-of-care assessments (speech audiometry). Additionally, by means of questionnaires we aim to determine the HCP's evaluation of the level of tolerance of the assessment for children, as well as the HCP's user experience and satisfaction with the CORGEE software. Finally, we will assess the usability of the CORGEE software as a counselling tool for persons with (suspected) hearing-related problems and their caregivers.

Inclusion Criteria:

1. Subcategory 1 (Infants & toddlers)

   • 0 - 2,5 years of age
   • Bilateral hearing loss OR
   • Suspicion of hearing loss (at least 1 criteria met): referred after FAIL on the new born hearing screening OR referred by the subject's paediatrician, general practitioner or otolaryngologist
   • Voluntary written informed consent of the legally authorised representative

2. Subcategory 2 (Kindergartners)

   • 2,5 - 6 years of age
   • Going to kindergarten
   • Bilateral hearing loss
   • Using hearing aids regularly for at least three months
   • Would be cooperative in behavioural testing
   • Knowledge of Dutch (at least 1 criteria met): Dutch as mother tongue OR enrolled in Dutch speaking education
   • Voluntary written informed consent of the legally authorised representative

3. Subcategory 3 (Primary school children)

   • 6 - 12 years of age
   • Going to primary school
   • Bilateral hearing loss
   • Using hearing aids regularly for at least three months
   • Would be cooperative in behavioural testing
   • Knowledge of Dutch (at least 1 criteria met): Dutch as mother tongue OR enrolled in Dutch speaking education
   • Voluntary written informed consent of the legally authorised representative

Exclusion Criteria:

• no exposure to Dutch
• known brain injuries
• acute illness