Predicting Pain and Disability From Voice Recordings in Low Back Pain

Chronic Low Back Pain (CLBP)

Chronic low back pain, Voice analysis, Oswestry Disability Index, Roland-Morris Questionnaire, Acoustic biomarkers, Pain assessment, VAS

This study aims to investigate whether pain intensity (assessed by Visual Analog Scale - VAS) and functional impairment (Roland-Morris Disability Questionnaire \[RMDQ\] or Oswestry Disability Index \[ODI\]) in patients with chronic low back pain can be estimated using acoustic features extracted from short voice recordings.

Chronic low back pain (CLBP) is one of the most common musculoskeletal complaints and a leading cause of disability worldwide. Assessing pain severity and its impact on function is primarily reliant on self-reported tools such as the Visual Analog Scale (VAS) and functional indices like the Oswestry Disability Index (ODI) or Roland-Morris Disability Questionnaire (RMDQ). However, these tools may be influenced by subjectivity and reporting bias. Therefore, objective and non-invasive methods for pain assessment are needed. This pilot observational study aims to investigate whether acoustic characteristics of voice recordings can be used to estimate pain intensity and functional impairment in individuals with CLBP. The hypothesis is that pain-induced changes in a patient's psychophysiological state may alter specific voice parameters (e.g., pitch, speech rate, pauses, intonation), which can be quantified through acoustic analysis. After obtaining ethical approval, data will be collected at the Department of Physical Medicine and Rehabilitation. A total of 110 participants aged 18-70 years with chronic low back pain will be recruited. Each participant will be asked to read a standardized one-minute text aloud, and their voice will be recorded in a controlled setting. On the same day, participants will complete the VAS and either the ODI or RMDQ questionnaires. Voice recordings will be analyzed using speech signal processing tools to extract features such as fundamental frequency, speech rate, articulation rate, pauses, and prosodic variations. These voice parameters will then be statistically correlated with VAS and disability scores. Participants will be grouped by their VAS score (0-10), with an equal number of participants targeted for each pain level to ensure balanced data distribution. Written and verbal informed consent will be obtained from all participants. All recordings and data will be anonymized, and participants will retain the right to withdraw from the study at any time without consequence. This study adheres to the principles of the Declaration of Helsinki and aims to provide early evidence supporting the feasibility of voice-based pain assessment. If successful, this approach could pave the way for low-cost, scalable, and objective tools to assist in the diagnosis and monitoring of chronic pain conditions.

No related papers found yet.