Nasality Evolution in a Nasal Polyposis Context : Multiparametric Evaluation : Articulatory (Imaging), Aerodynamics, Acoustics and Perception

"Nasal polyposis, a pathology related to speech Nasal polyposis (NP) is a chronic bilateral inflammation of the sinonasal mucosa characterized by the development of polyps in the sinus and nasal cavities. In the general population, its prevalence is 4% with a clear increase from the age of 50 years. Two clinical forms of NP exist: primary polyposis which is isolated or integrated in Widal's disease, secondary polyposis when it appears in the context of diseases such as cystic fibrosis, ciliary dyskinesia or immune deficiencies. NP is a diffuse rhinosinusitis, i.e. it affects both the anterior and posterior sinuses bilaterally. When drug treatment is not effective, and the patient's quality of life is impaired, surgery is proposed.

Surgical treatment consists of removing the polypoid mucosa to repermeabilize the nasal cavities and/or to enlarge the ostia (communication openings between the nasal cavities and the sinuses) and to allow drainage again. This surgery goes beyond the repermeation of the sinuses since it will allow to widen the nasal cavity with aerodynamic and acoustic consequences on speech; indeed, it will allow to bring a greater resonance of the vocal and nasal duct. To our knowledge, there are few studies that have focused on the impact of NP and its treatment on speech. Yet there is a real demand from patients to obtain answers related to the impact of this surgery on their voice. A pathology such as NP falls directly within the scope of clinical phonetics, i.e. the description and understanding of the mechanisms of articulatory and acoustic production of speech, applied to the pathology. The NP will disturb the communication on the production side by altering the nasal resonance essential in the production for the distinction between oral and nasal sounds but also on the receiver's side who will have difficulties to understand the message.

Thus, preoperatively and postoperatively, the resonant capacities will be disturbed: in the sense of a hyponasality before surgery due to the presence of polyps which will affect the quality of nasal sounds. Then potentially, the risk of a hypernasality after surgery because of the new anatomical configuration. Indeed, there is a diversity of acoustic effects that differ according to the sinuses involved, the nature and type of surgery and the anatomical and physiological specificities of the patient. The alteration of acoustic properties of phonemes after endonasal sinus surgery is diverse and not well known. It is known that NP surgery allows better communication between the sinus and nasal cavities but the real impact on nasal resonances is not known. The same is true for the activity of the vocal folds. There is a link between nasality and voicing, i.e. the vibration of the vocal folds necessary for the production of voiced sounds. Nasality favors voicing. Indeed, to establish voicing, the maintenance of an airflow through the glottic plane is necessary, a passage of air through the nasal cavity facilitates the maintenance of this flow. Preoperatively, since air cannot pass through the nasal cavity, compensatory mechanisms may appear, such as increased voicing or glottic thrust. Thanks to electroglottography, the quality of this voicing preoperatively and postoperatively will be observed.

The NP can therefore be an obstacle to the phenomenon of nasality. Indeed, the airflow in the nasal cavities will be disturbed because of the presence of polyps, which will have consequences on the resonance of these cavities as a whole. This project would provide a deeper understanding of the mechanism of nasal production. It results from the coupling of two cavities: the nasal and oral cavities, which will produce an anti-resonance phenomenon (a resonance phenomenon specific to nasality) and make this resonance process extremely complex. The impact of nasality on the acoustic signal is known (anti-formants etc.) but these manifestations remain diverse and dependent on individual anatomical variability. The impact of the sinuses on speech remains poorly understood today. The controversy between researchers on their influence on speech, encourages us to better explore their role in speech. But also to analyze the phenomenon of resonance before and after surgery to understand the consequences on speech intelligibility and the predictability of the consequences on the speech of patients postoperatively.

Understanding dysfunction in the context of NP will allow us to learn more about the functioning and production mechanisms of nasality in general.

The articulatory approach is central to our project. The articulation of speech differs according to the nature of the sound: the production of consonants will require a closure or a constriction in the oral cavity while for vowels the free passage of air is necessary. In addition to the mode of articulation, the sounds are also characterized by the cavity that enters in resonance: only the oral cavity or in addition to this last the nasal cavity. The resonance of the air within the nasal cavity will therefore allow the production of nasal sounds. Thus, the volume of the sinonasal cavities and the relationship between the nasal cavity and the sinuses have a role in the articulation of these nasal sounds. Medical imaging is our means of studying their articulation since it will allow us to quantify and visualize the sinonasal cavities and their mutual relationship. It will allow us to precisely grasp the extent of the impact of the sinuses on speech. These scannographic data will be compared with aerodynamic, acoustic and perceptual data.

The particularity of this pathology lies in the obstruction of the nasal and sinuses cavities by polyps. In the context of our study, it represents a model of disturbance of the nasality. Indeed, all the aspects of the nasality will be altered: the articulatory by the obstruction of the cavities, the aerodynamics by a disturbance of the circulation of the air flow within the nasal cavity, the acoustics by an alteration of the resonance of this flow, the vibration of the vocal folds by phenomena of compensation, and finally the perception of speech by others where the comprehension of speech is difficult. Thus, this dysfunction should to be observed in a multiparametric way in order to have a precise approach.

This population is thus ideal, in preoperative, it will allow to measure by aerodynamic, articulatory, acoustic, and perceptive data taking this dysfunction and to give precise answers. Then, postoperatively, these measurements will be repeated to observe a return to the expected functioning of the nasal cavity. Indeed, the cavities being no longer congested, a greater nasal airflow would be expected, which would have acoustic consequences on the resonance of nasal sounds. This could be accentuated because of the new anatomical configuration due to the surgery. Perceptually, the voice after surgery should no longer be considered as pathological. And on the glottic level, the phenomena of compensation would no longer be present.

In addition to the linguistics aspect, this population has the particularity of having a strongly impacted quality of life. the investigators would therefore like to measure this impact on quality of life before and after surgery.

The research focuses on a voice recording: an aerodynamic recording with a mask and an acoustic recording with a microphone and an electroglottograph.

Finally, with the acoustic data, a perception test will be designed to observe whether naive listeners perceive this alteration of nasality.

This research also focuses on self-assessment questionnaires of quality of life:

These questionnaires allow to understand the image that the speaker has of his/her voice. In the context of patients, the difficulties that people with rhino-sinus pathologies may have in their daily life will be understood.

The participants will complete two voice-specific questionnaires: the Voice Handicap Index (VHI) and the Speech Handicap Index (SHI).

Finally, a questionnaire of 4 questions dealing more specifically with the phenomena of nasality will be proposed. But also on the adaptations put in place by the speaker to speak and on the difficulty to read the corpus. One part of the questionnaire comes from a study by Behrman \& al (2002) and the other one was made by investigators.

The multiparametric approach is necessary for phonetic studies and even more so in the case of nasality, which is a complex phenomenon whose manifestations are different depending on whether one is looking at it from the articulatory, aerodynamic or acoustic point of view.

The manifestations of nasality are different depending on whether one is looking at it from the articulatory, aerodynamic or acoustic point of view. In order to have an optimal understanding of nasality and its alterations in the context of NP, the observation cannot be limited to aerodynamic, acoustic, and perceptual parameters alone.

To date, no speech-related study has sought to quantify sinus volume. The use of scannographic data (low-dose CT) is already used by the ENT surgeon to visualize the extent of polyps in the sinuses preoperatively and to verify the effectiveness of the intervention. It is essential to this study in order to theorize the morphological variability of the sinuses: bilateral volume asymmetry. Knowing the variability of the sinuses will allow to estimate their impact on nasal resonance and eventually on voice and speech. Postoperatively, the CT scan will allow to quantify the anatomical changes related to the surgery.

Imaging data, with these two scans before and after surgery, will allow to quantify the volumes of the nasal cavities and sinuses, and to quantify the volume of the polyps, which will lead to a 3D reconstruction. The latter will lead to acoustic simulations. The quantification of the volumes coupled with a 3D reconstruction will thus be central to this project and will allow to deepen the knowledge and to target the relationship between the sinuses and the resonance, and the relationship between the sinuses and the phenomenon of nasality.

The study of the impact of NP and surgery on speech could therefore represent a major help for the clinician to provide answers to patients and inform them of possible changes. Indeed, at present, there are no models capable of predicting the possible acoustic and aerodynamic consequences of surgery.

This articulatory aspect is particularly innovative because currently, to knowledge, there are no correlations between the medical imaging data of the sinuses (volume, geometry of the sinonasal cavities...) and the production of speech.

Patients will be recruited from the active file of patients followed in the recruiting center according to the eligibility criteria. After verification of the inclusion and non-inclusion criteria, the investigator will propose to the patients to participate in the study and will deliver the information and collect their non-opposition.

Then, the patient will be included in the study and an information and non-opposition form for the research will be given to him.

Clinical and para-clinical data (CT scan and nasal endoscopy), demographic and socio-demographic data, medical history, sinus quality of life questionnaire (SNOT-22) performed as part of the care will be collected in the observation booklet. In addition, voice recordings and voice and nasal-specific quality of life questionnaires (VHI, SHI and nasal-speech relationship) will be made for the purpose of the research. These data will be collected during two visits of the patient's care (preoperative visit and 3 months postoperative visit). The voice recordings will take place during these two visits and will last approximately 20 minutes. They will take place in the same way preoperatively and postoperatively.

Preoperative visit: The voice recording and the quality of life questionnaires at the preoperative visit will take place either during the visit with the ENT surgeon where the non-opposition will be collected, or during the anesthesia consultation. This will be determined according to the patient's expected waiting time during these appointments.

Post-operatively, the recording will take place during the ENT check-up appointment scheduled in the care pathway.

The recording will take place as follows:

• A presentation of the recording equipment by the research team.
• The patient will place the mask on his face and the electroglottograph around his neck.

During the recording, the patient will be asked to read a text.

• Finally, the patient will fill in the 3 voice quality questionnaires. Once the patients are recorded, a data processing phase will follow: segmentation of the recordings, analysis and acoustic and aerodynamic measurements.

Included patients will be reviewed at their M3 postoperative follow-up appointment with the same data collection and examinations as the preoperative visit.

The clinical data related to the current care and the data of the questionnaires will be collected in real time or a posteriori by the investigators and/or by the doctoral student Amélie Elmerich on a paper support at first. The aerodynamic and acoustic data will be collected on the doctoral student's professional computer simultaneously with the reading of the text by the patient, using the acoustic and aerodynamic data acquisition software QuickDaq.

The data directly collected in the CRF will be the recording data: aerodynamic, acoustic, electroglottographic recording, and the scores obtained from the questionnaires (VHI, SHI, nasal-speech questionnaire).

Care-related data:

This is a non-interventional study. All clinical, socio-demographic (age, sex, languages spoken, professional status), para-clinical, medical history data, the practice of an activity that requires an important use of the voice (singing, theater, wind instruments, teaching...), the consultation of a professional for voice, language, speech problems will be collected during usual care. All these data will be collected in a pseudonymized way (patient reference in the study: center number - enrollment number in the center - initials) from the patient's medical record (source record).

Socio-demographic data, the practice of an activity that requires an important use of the voice (singing, theater, wind instruments, teaching...) and the consultation of a professional for voice, language, speech problems. These data, specific to the individual, will allow us to understand certain results, they can have an influence on the results. They will also allow us to make hypotheses about a result that would be below or above the average values (flow, value of formants...). But also to observe through certain parameters which would be common (sex, age, profession etc.) between participants if there are similarities in their aerodynamic or acoustic results.

Research-related data:

Research-specific quality of life questionnaires will be collected in the Case Report Form (CRF).

The recordings (aerodynamic and acoustic data) will be collected by the doctoral student Amélie Elmerich at the Bicêtre hospital simultaneously with the recording of the patient, when he will read the text on the QuickDaq software.

The data will then be analyzed by her at the Phonetics and Phonology Laboratory (Sorbonne Nouvelle University).

The collection of voice data (acoustic and aerodynamic recording, quality of life questionnaires) will take place at the Bicêtre hospital. Then these data will be transferred to the Université Sorbonne Nouvelle where they will be processed by the doctoral student Amélie Elmerich under the responsibility of Lise Crevier-Buchman and Angélique Amelot. The statistical analysis will be carried out by the doctoral student Amélie Elmerich.

As this study is observational, there is no statistical justification for the number of patients. This number of participants (30) is chosen because it allows to make sufficient subgroups in terms of location of polyps and stage of severity in order to perform a phonetic analysis later on.

Inferential and descriptive statistics will be performed. The descriptive statistics will allow a first approach, it will be a question of using descriptive statistics in order to obtain average values, standard deviations and the range of duration of the sounds, oral and nasal airflow.

Inferential statistics will allow to complete the descriptive approach thanks to hypothesis tests (Anova for example) and thanks to the creation of a linear mixed model. The analysis will be conducted within each group (""Presence of polyps in the nasal and sinus cavities"", ""Presence of polyps in the sinus cavities"") by comparing different parameters (aerodynamic, acoustic, articulatory and perceptual) in order to highlight significant changes between preoperative and postoperative.

Then, variables such as gender or severity stage will be used to measure their potential impact on statistical results.

A linear mixed model will be established to observe the effect of surgery, gender and stage of severity on aerodynamic (oral and nasal airflow), acoustic (formant measurements) and articulatory (sino-nasal cavity volume) measurements.

This study is based on a voice recording. In case of missing data in a recording, the rest of the data present will still be taken into account and analyzed"