Longitudinal Ultrasound Kinematic Evaluation of Digastric Muscles

Background and technical information :

Ultrasound assessment of the anterior bellies of the digastric muscles (ABDM) has historically relied on a transverse (cross-sectional) approach placing the probe in the submental region. This conventional approach (CA) yields a non-interpretability rate of approximately 30% at rest and exceeding 50% during dynamic swallowing assessment, with poor reproducibility, preventing reliable access to functional muscle parameters such as thickness variation during contraction or pennation angle.

The proposed new ultrasound approach (NAEP) involves repositioning the probe to obtain a longitudinal (sagittal) section of the ABDM by rotating the probe 90° from the cross-sectional plane and applying a lateral displacement to align the muscle fibres along their axis. This approach is hypothesised to improve image interpretability and provide access to the pennation angle, a parameter not measurable with the CA.

All examinations are performed using a Sonoscape Expert 2 ultrasound scanner (7-10 MHz high-frequency linear probe), with the participant seated.

Study procedures Healthy volunteers: Each participant undergoes a single-session ultrasound assessment (≤45 minutes total) including image acquisition during four swallowing conditions - dry swallow and 5, 10, and 20 mL water boluses - in a randomised order (block randomisation, n=4). Both the NAEP and CA are applied to each condition, with the order of approaches also randomised at the subject level. Each bolus volume is assessed once per approach.

To assess reproducibility, a second complete examination is performed by a second independent operator (inter-rater reliability) and repeated by the first operator (intra-rater reliability), each separated by a standardised 5-minute interval.

Subjects with dysphagia: Ultrasound acquisition (NAEP and CA) is performed simultaneously during the clinically-indicated Volume-Viscosity Swallow Test (V-VST), following the standard V-VST volume sequence (non-randomised). No additional bolus volumes are administered beyond those clinically validated during the V-VST. The examination may be interrupted at any time based on clinical signs (coughing, wet voice, desaturation). The number of volumes assessed may therefore vary across subjects and will be systematically documented. Ultrasound image analysis is conducted offline, after the clinical assessment.

Parameters recorded

For each swallowing condition and each approach:

ABDM thickness at rest and during contraction (mm) Hyoid bone cranio-caudal displacement (mm) Pennation angle of the ABDM (NAEP only, when obtainable) An image is classified as interpretable if both superficial and deep muscular fasciae of the ABDM are identifiable over ≥50% of the visualised muscle length and allow reliable caliper placement for thickness measurement at rest and during contraction.

Blinded independent review All images are reviewed by a third independent assessor blinded to the primary operator's classifications, to assess inter-rater agreement on interpretability (Cohen's Kappa coefficient) and to conduct a pre-specified sensitivity analysis of the primary endpoint.

Statistical analysis The primary analysis uses a mixed-effects logistic regression model with feasibility (interpretable: yes/no, per swallow) as the dependent variable, approach (NAEP vs. CA), bolus volume, group (healthy/dysphagic), and order of administration as fixed effects, and a random subject intercept to account for repeated measures. The primary result is the odds ratio for the NAEP effect with 95% CI.

A supplementary analysis applies McNemar's test on subject-level overall feasibility (threshold: ≥75% interpretable swallows).

Secondary analyses include: ICC (two-way mixed model) for intra- and inter-rater reliability; linear mixed-effects models for the impact of bolus volume on ABDM thickness variation (Δ%) and hyoid displacement; Pearson/Spearman correlations between muscle parameters and hyoid movement; and between-group comparisons (healthy vs. dysphagic) using t-tests or Mann-Whitney tests with adjusted regression models.

The significance threshold is set at α=0.05 (two-sided) for the primary objective and α=0.01 for confirmatory secondary objectives.

Inclusion criteria

• For healthy volunteers:

• Aged 18 years or over
• Verbal, free, informed and express consent from the subject
• For subjects with swallowing disorder

• Subjects admitted to Forcilles Hospital to one of the short-stay or medical care and rehabilitation departments for diabetes and obesity, respiratory medicine, oncology, ENT and gastro-nutrition at Forcilles Hospital;
• Subjects who have undergone an assessment of swallowing disorders by a speech and language therapist;
• Subjects aged 18 or over at the time of inclusion;
• Enrolment in a social security scheme or eligibility for such a scheme;
• Verbal, free, informed and express consent from the patient.

Exclusion criteria

• For healthy volunteers :

• Under 18 years of age
• With a history of swallowing disorders
• With a medical history that could cause a swallowing disorder: stay in intensive care with orotracheal intubation; neurological or neuromuscular disease; cancer of the upper aerodigestive tract
• For Subjects with swallowing disorders

• Patients for whom there is no indication for swallowing tests;
• Individuals subject to a court-ordered protective measure;
• Patients under guardianship or administration;
• Patients subject to a care restriction order.