Scientific Overview

Disease monitoring in neuroscience is hampered by insensitive clinical scales which may not detect subtle changes in symptoms. By extracting over 550 features of speech and language, Winterlight Labs’ technology can measure fine-grained changes that traditional assessments cannot.

Our features and algorithms can be used in health research to help identify disease, track change over time, monitor disease severity, and detect response to treatment. Winterlight’s technology has been used to study neurodegenerative and psychiatric diseases and disorders, including Alzheimer’s Disease, Mild Cognitive Impairment, Frontotemporal Dementia, Parkinson’s Disease, Depression, Schizophrenia and more.

Speech in Alzheimer's Disease

Speech symptoms in Alzheimer’s disease include changes to the rate and rhythm of speech as well as its content and organization. Winterlight’s technology can be used to quantify these changes in order to detect signs of Alzheimer’s disease, monitor disease progression and measure response to treatments and therapies in clinical trials and other healthcare settings.

Recommended Speech Assessments

Comprehensive (up to every 6 months)

Paragraph reading
Picture description x2
Phonemic fluency
Semantic fluency
Journaling
Paragraph Recall

Brief (up to weekly, varying stimuli)

Picture description x2
Journaling

Speech Features Of Interest

Selected speech features that differ in Alzheimer’s disease:

Information content: Reduced number of correctly described items in picture description tasks
Coherence: Reduced semantic relatedness of descriptions to picture stimuli
Repetition: More similar or repetitive utterances
Use of prepositions: Reduced use of prepositions, which are relational words like "on", "over", or "under"
Use of pronouns: Increased use of pronouns, like "this", "that", or "it"

Select Publications

Generalizability and Robustness of Large Language Models Detecting Alzheimer’s Disease from Speech.

Novikova J. (2023)

GenBench at EMNLP (Empirical Methods in Natural Language Processing) International Conference.

Science

Scientific Overview

Speech in Alzheimer's Disease

Recommended Speech Assessments

Comprehensive (up to every 6 months)

Brief (up to weekly, varying stimuli)

Speech Features Of Interest

Select Publications

Generalizability and Robustness of Large Language Models Detecting Alzheimer’s Disease from Speech.

How Much Speech Data Is Needed for Tracking Language Change in Alzheimer’s Disease? A Comparison of Random Length, 5-Min, and 1-Min Spontaneous Speech Samples.

Robustness and generalizability of a speech based composite score for measuring disease progression in AD.

Automated detection of progressive speech changes in early Alzheimer's disease.

Speech phenotypes in cognitively healthy participants at risk of developing Alzheimer’s disease.

The generalizability of longitudinal changes in speech before Alzheimer’s Disease diagnosis.

Validation of an objective, speech-based object content score for measuring disease progression in AD.

Description of the Method for Evaluating Digital Endpoints in Alzheimer Disease Study: Protocol for an Exploratory, Cross-sectional Study.

Characterizing progressive speech changes in prodromal-to-mild Alzheimer’s disease using natural language processing.

Association Between Speech Characteristics And Cortical [18F]GTP1 Tau PET Tau Levels In Prodromal-To-Mild Alzheimer’s Disease

Using Digital Speech Assessments to Detect Early Signs of Cognitive Impairment.

Correlating natural language processing and automated speech analysis with clinician assessment to quantify speech-language changes in mild cognitive impairment and Alzheimer’s dementia.

Comparing Pre-trained and Feature-Based Models for Prediction of Alzheimer's Disease Based on Speech

Evaluation of speech-based digital biomarkers for Alzheimer’s Disease.

Linguistic features identify Alzheimer’s disease in narrative speech