Researchers at Penn State are pioneering a new approach to Alzheimer’s and dementia detection, harnessing the power of artificial intelligence to analyze speech patterns. With over 7 million Americans aged 65 and older currently suffering from Alzheimer’s disease—a 2025 figure from the Alzheimer’s Association—earlier diagnosis is critical for improved patient outcomes. Their work, recently published in the Journal of Alzheimer’s Disease Reports and Frontiers in Aging Neuroscience, proposes a framework that could detect cognitive decline years before traditional methods. “Traditional dementia screening tools are paper-based, subjective and resource-intensive…lacking sensitivity to subtle cognitive changes,” explains Hui Yang, Gary and Sheila Bello Chair in Industrial and Manufacturing Engineering at Penn State, highlighting the urgent need for scalable AI solutions in the face of a growing geriatric care crisis.
AI Speech Biomarkers Detect Early Neurodegenerative Disease
A novel approach leveraging artificial intelligence to detect early signs of neurodegenerative diseases like Alzheimer’s is under development at Penn State, potentially revolutionizing cognitive care. Researchers are focusing on speech as a rich source of biomarkers, analyzing patterns imperceptible to traditional methods. The new framework aims to address these limitations, offering an objective and non-invasive screening process completed “in under a minute.” This is particularly vital given the current shortage of geriatric specialists—roughly one geriatrician for every 10,000 geriatric patients—and high staff turnover in care facilities.
The team, including Kevin Mekulu, an industrial engineering doctoral candidate, is employing “agentic AI” – systems capable of independent reasoning and adaptation – rather than static models. “AI agents are not just scoring a test — they guide a screening interaction, adapt prompts based on a person’s responses,” Mekulu details. This dynamic interaction allows the AI to analyze complex linguistic features, including word choice and fluency, revealing cognitive decline years before conventional tests can. Ultimately, the goal is to transition cognitive care “from reactive to preventative.”
Agentic AI Enables Dynamic Cognitive Screening
A new paradigm in early detection of neurodegenerative diseases is emerging, leveraging the capabilities of “agentic” artificial intelligence to move beyond limitations of traditional cognitive assessments. Researchers are now focusing on AI’s potential to identify subtle changes indicative of conditions like dementia and Alzheimer’s, years before conventional methods can. The team’s innovation centers on analyzing speech patterns – a behavior Yang describes as “one of the most information-dense” humans produce – for nuanced linguistic biomarkers.
Researchers are also exploring the use of AI to analyze eye movements, physiological signals, and even motor behavior, envisioning a future where these systems “reduce administrative burden, highlight meaningful patterns and help transform cognitive care from reactive to preventative.” The team is currently validating these methods in assisted living and memory care environments, aiming to integrate them into real-world clinical practice.
Speech Pattern Analysis Reveals Subtle Linguistic Changes
A new approach to detecting early signs of neurodegenerative diseases centers on the nuanced analysis of human speech, potentially offering a significant leap forward in cognitive care. Researchers are leveraging artificial intelligence to identify subtle linguistic shifts often missed by traditional, paper-based screening methods. According to a 2025 report from the Alzheimer’s Association, over 7 million people aged 65 and older in the United States currently live with Alzheimer’s disease, highlighting the urgent need for earlier and more accurate diagnostic tools.
The team’s innovation focuses on “interpretable, speech-based biomarkers to capture subtle linguistic changes and cognitive decline years before traditional tools can,” completing screening in under a minute. The power of this method stems from the complexity of speech itself, which requires coordinated cognitive functions. “Speech is one of the most information-dense behaviors humans produce, requiring the coordination of memory, attention, language, executive function and motor planning,” Yang stated. The AI doesn’t simply score a test; it engages in a dynamic interaction, adapting prompts and integrating various signals to create a comprehensive assessment.
Speech is one of the most information-dense behaviors humans produce, requiring the coordination of memory, attention, language, executive function and motor planning – all cognitive systems that are affected early in neurodegenerative disease.
Yang
Expanding AI Analysis Beyond Speech to Holistic Signals
Current diagnostic methods for neurodegenerative diseases like Alzheimer’s often rely on subjective, paper-based assessments that lack sensitivity to early cognitive shifts, a problem researchers at Penn State are addressing with advanced artificial intelligence. Beyond simply scoring tests, their work focuses on “agentic AI” – systems capable of dynamic interaction and independent task execution, offering a significant leap forward. According to Kevin Mekulu, this transforms screening “from a one-time measurement to an evolving process that better reflects how cognitive decline occurs in patients.” This contrasts with existing “static” AI models used in healthcare, which offer limited adaptability.
The team’s innovative approach isn’t solely focused on speech, however, but envisions a broader analysis of patient signals. “Interpreting all these signals together offers clinicians a more holistic view,” Mekulu adds, moving beyond simple pass/fail results. Currently, the U.S. faces a critical shortage – roughly one geriatrician for every 10,000 patients – making scalable AI solutions increasingly urgent.
