summary: By analyzing human speech, artificial intelligence can detect signs of mild cognitive decline and Alzheimer’s disease, even when there are no obvious symptoms. This technique can be used as a simple screening method to identify early signs of cognitive impairment.
sauce: UT Southwestern
According to researchers at UT Southwestern Medical Center, who led the study in the Alzheimer’s Association publication, new technology that can capture subtle changes in a patient’s voice could help doctors identify cognitive impairment before symptoms begin to appear. It may help diagnose Alzheimer’s disease. Diagnosis, evaluation and disease monitoring.
“Our focus was to identify subtle language and voice changes that are present in the very early stages of Alzheimer’s disease but that are not readily recognizable by family members and personal physicians,” said Peter of UT Southwestern’s. said Ihab Hajjar, MD, professor of neurology at O’Donnell Jr. Brain Institute.
Researchers used advanced machine learning and natural language processing (NLP) tools to evaluate speech patterns of 206 people. 114 met the criteria for mild cognitive decline and 92 had no impairment. The team then mapped these findings to commonly used biomarkers to determine their effectiveness in measuring disability.
Research participants enrolled in a research program at Emory University in Atlanta underwent several standard cognitive assessments before being asked to record spontaneous 1- to 2-minute descriptions of works of art. .
“Recorded photographic descriptions provide an approximation of speech ability that can be studied via artificial intelligence to determine vocal motor control, density of ideas, grammatical complexity, and other speech features.” ,” said Dr. Hajjar.
The research team compared the participants’ voice analysis with cerebrospinal fluid samples and MRI scans to determine how accurately the digital voice biomarkers detected both mild cognitive impairment and Alzheimer’s disease status and progression. bottom.
“Before the development of machine learning and NLP, a detailed study of a patient’s speech patterns was very labor intensive and often unsuccessful because early-stage changes were often undetectable by the human ear. We did,” said Dr. Hajjar.
“This new test method has been successful in detecting people with mild cognitive impairment, and more specifically in identifying patients with evidence of Alzheimer’s disease, even if they cannot be easily detected using standard cognitive assessments. It worked.”
During the study, researchers took less than 10 minutes to capture the patient’s voice recording. Conventional neuropsychological testing typically takes several hours to administer.
“If confirmed by large-scale studies, using artificial intelligence and machine learning to study voice recordings has the potential to provide primary care providers with an easy-to-implement screening tool for at-risk individuals. Yes,” said Dr. Hajjar. “Earlier diagnosis gives patients and families more time to plan for the future and gives clinicians more flexibility to suggest promising lifestyle interventions.”
Before joining UTSW in 2022, Dr. Hajjar worked with a team of researchers at Emory, who was director of the Clinical Trials Unit at the Goizueta Alzheimer’s Disease Research Center. of a follow-up study at UTSW funded by a National Institutes of Health grant.
Funding: Work in this study was supported by grants from the National Institutes of Health/National Institute on Aging (AG051633, AG057470-01, AG042127) and the Alzheimer’s Disease Drug Discovery Foundation (20150603).
Dr. Hajjar holds the Pogue Family Distinguished University Chair in Alzheimer’s Clinical Research and Care in memory of Maurine and David Weigers McMullan.
About this AI and Alzheimer’s research news
author: press office
sauce: UT Southwestern
contact: Press Office – UT Southwestern
image: image is public domain
Original research: closed access.
“The development of digital voice biomarkers and their association with cognition, cerebrospinal biomarkers, and neural representation in early Alzheimer’s disease,” Ihab Hajjar et al. Alzheimer’s disease and dementia: diagnosis, evaluation, and disease monitoring
overview
Development of digital voice biomarkers and their association with cognition, cerebrospinal biomarkers, and neural representation in early Alzheimer’s disease
prologue
Advances in natural language processing (NLP), speech recognition, and machine learning (ML) have made it possible to investigate language and acoustic changes that were previously difficult to measure. Digital speech bio-biology of Alzheimer’s disease (AD) He developed a process for deriving lexical and acoustic measures as markers.
method
From 92 participants without cognitive impairment (40 Aβ+) and 114 participants with cognitive impairment (63 Aβ+), connected speech, neuropsychological, neuroimaging, and cerebrospinal fluid (CSF) AD We collected biomarker data. Acoustic and lexical semantic features were derived from audio recordings using an ML approach.
result
Lexical meaning (area under the curve [AUC] = 0.80) and acoustic (AUC = 0.77) scores showed higher diagnostic performance for detecting MCI compared to the Boston Naming Test (AUC = 0.66). Only the lexical semantic score detected Aβ status (p = 0.0003). Acoustic scores related to hippocampal volume (p = 0.017), while the lexical-semantic score associated with CSF amyloid-β (p = 0.007). Both measures were significantly associated with disease progression over 2 years.
discussion
These preliminary findings suggest that derived digital biomarkers may identify cognitive deficits in preclinical and prodromal AD and predict disease progression.
