Speech and Digital Biomarkers for Parkinson's Disease: From Multimodal Sensing to Early Detection and Disease Monitoring

Parkinson's disease (PD) is a progressive neurodegenerative disorder characterised by heterogeneous motor and non-motor symptoms that can emerge years before clinical diagnosis. Current assessment relies heavily on episodic clinical evaluations and subjective rating scales, which may fail to capture subtle, continuous, or real-world changes in disease state. Among digital sensing modalities, changes in speech, voice production, and language function, such as hypokinetic dysarthria, acoustic degradation, articulatory imprecision, and altered syntactic complexity, are some of the earliest and most sensitive indicators of underlying neurodegeneration. Advances in digital health technologies now offer opportunities to capture these objective speech and vocal markers alongside complementary multimodal signals which can be captured by smartphones, wearable, and ambient intelligence devices, including gait, sleep patterns, EEG, among other combined with artificial intelligence and multimodal data integration, these technologies could support earlier detection, longitudinal monitoring, treatment-response assessment, and personalised disease management beyond specialist clinical settings.Despite rapid advances in digital health and artificial intelligence, the translation of digital phenotyping techniques and biomarkers into clinical practice remains fragmented. Many studies focus on isolated signals in controlled laboratory environments, leaving substantial challenges in multimodal fusion, longitudinal validation, acoustic interpretability, reproducibility, and cross-population generalisability unresolved. This Research Topic aims to bring together interdisciplinary research that advances robust, clinically meaningful digital biomarkers across the Parkinson’s disease continuum, placing particular emphasis on vocal, linguistic, andsensor-based phenotypes. We welcome studies evaluating how speech and language changes correlate with neuropathology and motor progression, developing validated AI-driven analytics, and establishing how acoustic markers integrate with broader digital health data in real-world settings. By bridging linguistic and speech sciences, digital health, and neurology, this collection seeks to establish methodological best practices and accelerate clinical translation into personalised Parkinson’s disease care.This Research Topic welcomes interdisciplinary contributions advancing digital and speech-based biomarkers for Parkinson’s disease across screening, diagnosis, monitoring, prognosis, and treatment-response evaluation. We particularly encourage submissions exploring:* Speech and Vocal Biomarkers: Acoustic, phonatory, articulatory, and prosodic analysis for prodromal detection and symptom tracking;* Linguistic and Cognitive-Language Markers: Computational linguistics, natural language processing (NLP), semantic and syntactic complexity, and conversational dynamics reflecting cognitive or motor decline;* Multimodal Integration: Combining speech and audio signals with gait, tremor, wearable kinematics, smartphone sensing, sleep, EEG, and other physiological and behavioural signals;* Neurobiological Underpinnings: Linking acoustic and digital alterations to neural circuitry, structural and functional neuroimaging, electrophysiology, sleep and EEG biomarkers, and disease progression;* Machine Learning and AI Methodologies: Explainable AI, robust feature extraction, multimodal data integration, and AI-based approaches for biomarker discovery and validation, including the use of foundation models;* Clinical Translation and Implementation: Longitudinal validation, wearable and smartphone sensing, remote and passive monitoring, ecological validity, device interoperability, data privacy, and ethical frameworks.We welcome Original Research, Methods, Brief Research Reports, Reviews, Systematic Reviews, Perspectives, and Technology and Code articles that connect computational and technological innovation with speech science and clinical applications in Parkinson’s disease. Topic editors: Fasih Haider, Saturnino Luz, Biao Zeng Source: https://www.frontiersin.org/research-topics/86631/speech-and-digital-biomarkers-for-parkinsons-disease-from-multimodal-sensing-to-early-detection-and-disease-monitoring