Multimodal paradigm combining surface electromyography (sEMG), kinematic tracking, and acoustic techniques to measure craniofacial muscle activities, including temporalis, masseter, and submental complex, the motions of speech organs, such as the jaw and lips, and the acoustic features of speech.

Research Projects

The Speech Science and Disorders (SSD) Lab uses multimodal techniques, including craniofacial surface electromyography (sEMG), kinematic tracking, and acoustic recording, to develop personalized diagnostic, prognostic, and management models of progressive communication disorders secondary to neurodegenerative diseases.

Objective assessment and phenotyping

This project aims to develop a clinically translatable noninvasive multimodal tool, combining craniofacial sEMG and acoustic recording with novel automated computational techniques, for automated objective assessment and phenotyping of progressive communication disorders in individuals with neurodegenerative diseases. Different neurodegenerative disease models with distinct pathophysiology, such as amyotrophic lateral sclerosis (ALS) and  Parkinson’s disease (PD), have been tested, generating a rich data set for validating the utility of the tool in detecting, disentangling, and phenotyping the highly heterogeneous communicative changes across the motor and cognitive-linguistic domains. The outcomes of these studies improve the definition and stratification of individuals with neurodegenerative diseases, providing foundational information to guide clinicians in selecting the right intervention for the right person at the right time, thereby facilitating patient-centered care for progressive communication disorders.

Subclinical marker development

This project focuses on identifying novel objective markers to detect and quantify clinically indiscernible subclinical changes in speech-related craniofacial neuromuscular performance during the preclinical/prodromal phase of neurodegeneration. In collaboration with Dr. Marziye Eshghi's lab at the Massachusetts General Hospital, we have identified a variety of changes in mandibular and facial muscles in asymptomatic individuals who are genetically at risk for developing Alzheimer's disease (AD). These subclinical changes, following a variable period of progression, are highly likely to culminate in clinical symptoms and functional decline, ultimately contributing to the clinical diagnosis of AD. Therefore, detecting these subclinical changes can significantly improve the screening, diagnosis, monitoring, and development of tailored therapeutic interventions for individuals with AD.

Rhythm-based speech assessment and management models

Speech is hierarchically structured as a cascade of oscillations, reflecting multiscale rhythmic modulation of linguistic context at the delta (0.9-2.5 Hz), theta (2.5-12 Hz), and beta/gamma (12-40 Hz) timescales. These timescales correspond to the rhythms of prosodic, syllabic, and sub-syllabic units, collectively constituting a multiscale rhythmic template to entrain both the speech production and perception systems. Such a multiscale rhythmic template plays a central role in the effective transmission and processing of linguistic information for functional communication. Building upon this concept, we have developed an explanatory model that links the production and perception aspects of speech communication through a key mediator—multiscale rhythmic modulation—to inform rhythm-centered assessment and management paradigms for progressive communication disorders.

Based on the explanatory model, we have carried out two lines of research. The first line of work is focused on developing and validating an objective measurement tool to detect and characterize rhythmic disturbances in both articulatory and acoustic activities across individuals with different neurogenerative diseases. The second line of work integrates a computational speech synthesis model with a perceptual study to (1) simulate tailored intervention strategies based on individual rhythmic characteristics and (2) evaluate the perceptual outcomes of the resulting synthetic speech stimuli.

The above two lines of research provide complementary insights to inform rhythm-centered models for objective assessment and tailored management of rhythmic disturbances, paving the way for a new direction in personalized care for progressive communication disorders. This contribution aligns well with the growing emphasis on precision medicine and measurement-based care in healthcare professions.