• Citer

Spectro-Temporal Study of the Frequency-Following Response: A new technique to study the temporal neural code in the auditory pathway

par Díaz Cortés, Macarena
Président du jury Kinnaert, Michel
Promoteur Nonclercq, Antoine
Co-Promoteur Deltenre, Paul
Publication Non publié, 2025-01-30

Thèse de doctorat

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

Hearing loss results from a bewildering number of various etiologies that can be congenital or acquired, with complex underlying mechanisms, which have been revealed by advances in genetic diagnosis. Because of this diversity and the critical relevance of early detection of hearing loss, followed by early appropriate compensation targeted to the implied defect, up-to-date pediatric audiology requires adherence to the Cross-Check Principle. This principle uses multiple tests with differential sensitivity to improve diagnostic accuracy. The frequency-following response (FFR) has emerged as a valuable tool reflecting the physiological mechanism of phase-locking along the auditory pathway. Phase-locking is essential for precise auditory processing, such as speech understanding in noise.Complex stimuli (such as a pair of two tones with primaries f1 and f2) evoke FFRs that capture phase-locking to both the stimulus envelope (envelope-following response, EFR), at f2-f1 and its harmonics, as 2(f2-f1), and the temporal fine structure (TFS), including f1, f2 and 2f1-f2. The latest frequency component is the cubic difference tone (FFR-CDT). Different sources, ranging from the cochlea to the cortex, can contribute to the FFR. Although the EFR is traditionally viewed as neural in origin, near-field recordings in animal models and invasive human measurements have revealed pre-neural components. These components require further exploration to better evaluate their potential contribution to clinically relevant far-field scalp recordings. The origins of the FFR-CDT remain uncertain, with prior research primarily examining its amplitude only.This thesis aims to enhance our understanding of the sources contributing to the FFR in realistic clinical recording conditions for difficult-to-test sleeping children and to assess its potential as a clinical tool to refine the diagnosis of pathologies. By applying the generalized Primary Tone Phase Variation Method, spectral components of the FFR were isolated, enabling a detailed spectro-temporal analysis. Recordings from horizontal (H, earlobe-to-earlobe) and vertical (V, vertex-to-neck) channels were used to evaluate putative peripheral and central contributions.The FFR-CDT was recorded at nine CDT frequencies (fCDT) from 79 controls, and at fCDT = 442 Hz from 26 subjects with auditory neuropathy spectrum disorder (ANSD) and four with brainstem dysfunction. Their onset latency and phase-locking values (PLVs) were evaluated across frequencies. The CDT-H showed a constant PLV across fCDT, and latencies within cochlear traveling wave delay estimations, suggesting a cochlear origin. The CDT-H presence in cases of cochlear nerve deficiency and genetically diagnosed ANSD supported the pre-neural contribution. In contrast, the CDT-V showed a low-pass PLV transfer function and longer onset latencies than CDT-H, indicating a possible cochlear nucleus origin. Furthermore, the absence of CDT-V in most ANSD cases and all subjects with brainstem dysfunction pointed to a central origin. These findings highlight the separation of peripheral and central sources contributing to the FFR-CDT.Normal EFR-H recordings from two cases with severe brainstem dysfunction but normal cochlear and cochlear nerve function further confirmed the hypothesis of its cochlear nerve origin in sleeping children. On the other hand, the EFR-H was detected in five out of 26 patients with ANSD who had diminished or absent neural responses but preserved cochlear potentials. The EFR-H had shorter latencies, shorter response durations, and lower PLVs than the controls, suggesting a pre-neural origin, with the asymmetrical mechanoelectrical transduction process at the sensory hair cells as the main contributor. Three genetically diagnosed ANSD cases with absent post-synaptic components did not produce recordable EFRs, possibly due to limitations in our recording method or congenital defects in the inner hair cells' mechanoelectrical transduction reducing or abolishing the pre-neural signal. The EFR-V was absent in all ANSD subjects except in two with milder neural desynchronization, indicating that a central source was sufficiently activated in these cases despite peripheral dys-synchrony.The analysis of EFR, CDT, and the responses to the primaries: F1 and F2, provided insights into the physiopathology of ANSD. Lower PLVs in the contralateral ears of unilateral ANSD cases revealed subtle deficits missed by standard audiological tests. Moreover, F1 and F2 analysis showed their potential for assessing outer hair cell function at different tonotopic locations. This is particularly interesting in ANSD children for whom cochlear assessment is often compromised by conductive hearing loss and the frequency-limited assessment offered by the click-evoked cochlear microphonic. Additionally, the study of the EFR, click-evoked auditory brainstem responses, and auditory steady-state responses (ASSR) in children with severe brainstem dysfunction emphasized the differential sensitivity of these techniques. While the ASSR was blind to brainstem dysfunction, the EFR revealed preserved central responses despite abnormal latencies and PLVs, emphasizing the importance of using multiple methods for accurate auditory assessment.This research contributed to a better understanding of the FFR-CDT and EFR sources, improving the ability to locate functional defects along the auditory processing chain. It also provided compelling evidence of the potential of the FFR as a clinical tool to refine the diagnosis of brainstem dysfunction and ANSD. The comprehensive parametric analysis of the FFR could eventually contribute to the early identification and intervention of pathologies, such as genetically defined ones. In this context, it may serve as a post-treatment evaluation tool due to its unique sensitivity to sound time-encoding compared to standard audiometric techniques.