Digital speech and language markers of extrasylvian dysfunction in neurodegeneration

Aging-related neurodegenerative disorders (NDs), including Alzheimer’s disease (AD), Parkinson’s disease (PD), and behavioral variant frontotemporal dementia (bvFTD), are a central public health concern. These incurable diseases are disabling, deadly, costly, and increasingly prevalent in Chile and beyond. Early detection is vital to implement timely neuroprotective habits, optimize treatments, and reduce related expenses. Worryingly, standard diagnostic tools (clinical tests, brain scans, biofluid markers) are expensive, stressful, subject to scheduling delays, and unavailable in many cities and regions. A fruitful alternative consists in establishing digital speech and language (DSL) markers –i.e., acoustic and linguistic features derived from speech to identify and phenotype NDs. Due to its cost-effectiveness, non invasiveness, time-efficiency, remote applicability, and scalability, this approach is being leveraged in research projects, diagnostic settings, and clinical trials. Yet, mainstream DSL research targets generic speech/language functions through stock metrics that were not designed for these NDs. Such generic functions are grounded in perisylvian brain regions. These cortical areas are affected in language-dominant syndromes (e.g., aphasia), but they are fully or partly spared in early stages of AD, PD, and bvFTD. Indeed, these three NDs mainly affect extrasylvian neurocognitive systems (temporo-hippocampal regions subserving semantic memory in AD, frontostriatal circuits underpinning motor functions in PD, and orbitofrontal-insular areas mediating socio-emotional behavior in bvFTD). Unsurprisingly, existing studies are inconsistent in their capacity to identify syndromes, sometimes offering good results but often yielding large error margins and failing to capture diagnostically relevant brain alterations. Also, the field lacks multimodal neural validation and neglects Latino populations despite their heightened risk for NDs. Thus, the need arises for new DSL frameworks strategically designed to capture these NDs’ core deficits and neural disruptions, especially in Latinos. Such gaps can be bridged through a transdisciplinary effort focused on Chilean participants, integrating language science and behavioral neurology with the fields of audio engineering, computer science, artificial intelligence, and multimodal neuroscientific methods [magnetic resonance imaging (MRI), functional MRI (fMRI) and electroencephalography (EEG)], and machine learning analyses.