Dysphonia Severity Index

Abstract

The Dysphonia Severity Index (DSI) is a specialized, objective measure developed for the quantitative assessment of voice quality, particularly in relation to vocal disorders like dysphonia. It employs a multivariable statistical approach, combining several weighted acoustic and aerodynamic parameters into a single index score. This composite score is designed to optimally differentiate various voice types and accurately reflect the severity of vocal impairment. The DSI provides a standardized, objective alternative or complement to subjective perceptual assessment methods.

Keywords

Dysphonia Severity Index, DSI, Dysphonia, Voice Assessment, Acoustic Analysis, Maximum Phonation Time, Jitter, GRBAS scale, Vocal Function, Objective Measurement.

Authors

Wuyts FL, de Bodt MS (1996)

Purpose

The primary purpose of the DSI is to provide a comprehensive, quantitative, and objective measurement of voice quality. Prior to its development, the assessment of vocal pathology relied heavily on subjective perceptual scales, such as the GRBAS scale. The DSI was engineered to reduce the inherent variability and subjectivity associated with human perception by translating vocal characteristics into a single, reliable numerical value, aiding in clinical diagnosis and tracking treatment efficacy.

The index serves as a powerful tool in clinical phoniatrics and laryngology for documenting the severity of vocal impairment. By integrating diverse physiological and acoustic parameters, the DSI aims to capture the multidimensional nature of dysphonia in a standardized manner, facilitating cross-clinic comparisons and research efforts. The methodology ensures that the assessment of stem en spraakfuncties (voice and speech functions) related to the hoofd/hals (head/neck) region is highly standardized.

Construct

The DSI measures the construct of vocal integrity and the severity of dysphonia. It is fundamentally based on the concept that vocal impairment can be accurately modeled by combining key acoustic and aerodynamic indicators that reflect underlying laryngeal function and vocal fold vibration regularity. The scale is based on principles established by the subjective GRBAS scale (developed by Hirano in 1981), translating perceptual attributes (Grade, Roughness, Breathiness, Asthenia, Strain) into quantifiable physical metrics.

The index calculation involves four specific weighted variables, each selected because it is a distinct indicator of vocal function: the Maximum Phonation Time (MPT), the highest frequency achieved (F0 max), the lowest intensity produced (I min), and the short-term perturbation measure known as Jitter. These variables are statistically weighted to optimize the discrimination between normal and dysphonic voices, resulting in a score that theoretically ranges from +5 (indicating a healthy, strong voice) to -5 (indicating severe dysphonia).

Validity

Although specific validity statistics (like criterion or concurrent validity) were not detailed in the original source, the DSI’s validity is generally established by its strong correlation with established subjective measures, particularly the Grade (G) component of the GRBAS scale. Initial validation studies by Wuyts and colleagues demonstrated that the DSI score accurately reflects the perceptual severity of dysphonia, confirming its clinical utility and relevance.

The construct validity is supported by the selection of variables—MPT, F0 max, I min, and Jitter—which are widely accepted in voice science as fundamental indicators of respiratory support, laryngeal muscular strength, and vocal fold regularity. The multivariable statistical model ensures that the combined score represents the global severity of vocal impairment better than any single parameter alone, providing a robust measure for objective assessment.

Reliability

The DSI is inherently highly reliable due to its reliance on instrumental and computer-assisted measurements, which significantly minimizes the inter-rater variability common in subjective perceptual assessments. The reliability of the DSI is primarily dependent on the quality of the acoustic recording equipment and the precision of the software used for calculation (e.g., the Multidimensional Voice Program).

Measures of test-retest reliability typically show high consistency, provided the patient’s vocal condition remains stable between measurement periods. The objectivity provided by the DSI addresses a major limitation of traditional voice assessment, contributing significantly to the scale’s overall reliability in both clinical practice and research settings. The use of standardized procedures for vocal tasks further enhances the replicability of the DSI score.

Factor Analysis

Since the Dysphonia Severity Index is a composite score derived from four predetermined and weighted variables, it does not typically undergo traditional exploratory factor analysis in the same manner as a latent construct psychological questionnaire. Instead, the index’s structure was determined through statistical modeling (likely multiple regression or discriminant analysis) during its development to find the optimal weightings of the four parameters that maximized the prediction of perceived dysphonia severity (the ‘G’ score of GRBAS).

The four parameters—Maximum Phonation Time, highest frequency, lowest intensity, and Jitter—are treated as distinct, yet complementary, factors contributing to the overall construct of vocal integrity. The final DSI calculation formula mathematically represents this established factor structure, where each variable contributes proportionally to the final index score.

Instrument

Test Type: Instrumental / Objective Measurement

Format: Computerized acoustic and aerodynamic analysis based on standardized vocal tasks.

Language Available: Not applicable for the primary measurement variables, as the parameters (frequency, time, intensity, perturbation) are acoustic metrics. Instructions for the patient are typically language-dependent.

Population Group: Clinical and General Population

Age Group: Adults

Population Details: Individuals presenting with vocal complaints, suspected or diagnosed dysphonia, or those undergoing treatment for voice disorders. The scale specifically targets stem en spraakfuncties (voice and speech functions) related to the hoofd/hals (head/neck) region.

Test Methodology: Patients perform specific vocal tasks, such as sustaining a vowel sound at comfortable pitch and loudness, and producing the maximum range of pitch and loudness. Specialized computer software, such as the Multidimensional Voice Program (MVP) by Kay Elemetrics (now PENTAX Medical), captures and analyzes the acoustic signal to calculate the four required variables. These variables are then input into the weighted DSI formula to generate the final score. The original PDF documentation for the instrument can be downloaded here: Toelichtingsformulier.

Keywords

Vocal Assessment, Phonation, Laryngology, Speech Pathology, Maximum Phonation Time, F0 maximum, Intensity minimum, Jitter, Objective Voice Index, Kay Elemetrics.

Authors

Author ORCID Identifier: Information not provided in source.

Affiliation Email addresses: Information not provided in source.

Correspondence Address: Information not provided in source.

Permissions & Fee and Test Year

The Dysphonia Severity Index was introduced by Wuyts FL and de Bodt MS in 1996. Calculation of the index relies heavily on proprietary software, specifically programs like the Multidimensional Voice Program (MVP) of Kay Elemetrics. Therefore, access and usage often require the purchase or license of the necessary hardware and software required for accurate acoustic analysis, which may involve associated fees.

Reference’s

• Wuyts FL, De Bodt MS, Molenberghs G, Remacle M, Heylen L, Nieboer K, Van de Heyning PH. (1996). The Dysphonia Severity Index: an objective measure of vocal quality based on a multiparameter approach. Journal of Speech and Hearing Research.
• Hirano M. (1981). Clinical examination of voice. Springer-Verlag. (Reference for the GRBAS scale foundation).

Items of the Dysphonia Severity Index

IMPORTANT: The following scale items must be preserved in their original language and must not be changed in any way.

The DSI is a calculated index based on four instrumental variables:

• maximale fonatietijd
• hoogste frequentie
• laagste intensiteit
• jitter