The Laryngovibrogram as a normalized spatiotemporal representation of vocal fold dynamics.

Laryngeal high-speed video (HSV)-endoscopy allows for fast, non-invasive diagnosis of voice disorders and forms the basis for a comprehensive quantitative analysis of the vocal folds' (VFs') spatiotemporal vibrational behavior. Previous approaches, such as the Phonovibrogram (PVG), describe the vibrational behavior of vocal folds (VFs) based exclusively on the time-varying glottal opening. However, focusing solely on the glottal area overlooks the full extent and dynamic behavior of the VF tissue, factors that are crucial for the voice production process. This complicates clinical interpretation and, thus, the comparability of vibrational dynamics in both cross-sectional and longitudinal interventional studies. To address these limitations, this work aims to extend the PVG to provide a more comprehensive representation of the vibrational behavior across the entire VF tissue. Here, we present the Laryngovibrogram (LVG), which is obtained by segmenting not only the glottal area but also the VFs' tissue, providing a compact quantitative representation of the VFs' vibrational behavior. The potential of the proposed LVG representation was investigated on 73 HSV recordings from healthy (38 HSVs) and pathological subjects (35 HSVs) in stationary as well as non-stationary phonations. It is demonstrated that the LVG reliably maps the vibrational behavior along the entire length of the VFs tissue for both physiological and pathological phonations. Compared to PVG-based measures, LVG-based measures exhibited greater stability in healthy subjects, allowing for a narrower normative range, and showed stronger effect sizes in differentiating clinical groups, suggesting a more robust assessment of vibratory impairments. By scaling the vibration amplitude relative to the length of the segmented VF tissue, the VF vibrations are normalized, enabling meaningful quantitative intra- and inter-individual comparisons. Additionally, calculating the angle enclosed by the two VFs makes it possible to analyze transient effects that occur during non-stationary phonation maneuvers, such as voice onset. By integrating information about the VF tissue, the LVG introduced here represents a paradigm shift in the analysis of laryngeal dynamics from focusing solely on the glottal area to a holistic analysis of the entire VF kinematics, which might improve pathology detection accuracy, reduce subjective assessment errors, and optimize treatment follow-ups, ultimately enhancing both clinical diagnostics and therapeutic outcomes.