CYMO Use Case
Empowering Developers, Researchers, and Data Scientists in Building XAI-Powered Models
Digital Biomarkers for Brain and Mental Health
Unlock the potential of CYMO to create advanced, explainable AI models that harness digital biomarkers for improved brain and mental health diagnostics and monitoring.
Extract Digital Biomarkers with CYMO in Action
Uncover Subtle Signals
Uncover subtle statistical signals in verbal behavior from speech or text associated with neurodegenerative and psychiatric disorders
Trust in Rigorous Validation
Benefit from digital biomarkers that undergo rigorous computational validation, ensuring reliable association with pathogenic states.
Enhance Classification and Predictive Models
Build advanced predictive models for early and accurate detection of disorders, trained on CYMO-powered digital biomarkers.
Monitor and Track Disease Progression
Harness CYMO's powerful features to accurately monitor and track disease progression over time.
some background
Digital biomarkers are set to revolutionize healthcare, driving groundbreaking advancements in disease prevention, diagnostics, treatment, and monitoring.
DEFINITION OF A DIGITAL BIOMARKER
A digital biomarker is an objective, quantifiable measure of physiology and/or behavior, serving as an indicator of biological, pathological processes, or responses to exposure or intervention, derived from digital measurements and subjected to rigorous computational validations.
To dive deeper into this topic, explore our scientific research publications.
Why Language-Based Digital Biomarkers?
Language-based digital biomarkers enable objective, scalable, ecologically valid, and long-term monitoring through frequent or continuous assessment that is minimally obtrusive and can operate in the background of everyday activities.
In contrast to blood and imaging biomarkers that measure elevated levels of tau and amyloid proteins, language-based digital biomarkers are highly sensitive to early cognitive changes, detecting these alterations long before structural changes become apparent on neuroimaging.
Language impairments and dysfunctions exhibit unique patterns across diverse disorders, aiding in differential diagnosis and offering a clearer understanding of an individual’s mental health status.
Comparison of Language-Based Digital Biomarkers:
This table evaluates language-based digital biomarkers derived from spontaneous speech and writing against blood and imaging biomarkers and clinical rating scales used in neurodegenerative and psychiatric diagnostics.
| Language-Based Digital Biomarkers | Blood Biomarkers | Imaging Biomarkers | Clinical (Rating) Scale | Edit | |
|---|---|---|---|---|---|
| Scalable | ○○○○○ | ○○ | ○ | ○ | |
| Cost-effective | ○○○○○ | ○○ | ○ | ○○ | |
| Non-obstrusive | ○○○○○ | ○○○ | ○ | ○○○ | |
| Patient compliance | ○○○○○ | ○○ | ○ | ○○○ | |
| Predictiveness | ○○○○○ | ○○○ | ○○○ | ○○ |
Why Digital Biomarkers by Exaia?
Rigorous Scientific Validation
Exaia's digital biomarkers are validated through peer-reviewed research, demonstrating strong predictive power for diverse neurodegenerative and psychiatric disorders
Superior Performance
The ML models consistently outperform benchmarks and excel in international shared task competitions, showcasing their reliability and effectiveness.
Detection of Alzheimer's Dementia
F1 score of 83.07% at 2021 ADReSSo challenge organized by Carnegie Mellon University.
Detection of Chronic Stress
F1 75.8% at SMM4H Shared Task 2022 organized by Roche Innovation Center & Cedars-Sinai Medical Center.
Detection of Social Anxiety
F1 83.8% at SMM4H Shared Task 2023 organized by Roche Innovation Center & University of Pennsylvania at AMIA2023.
End-to-End Platform
Exaia is the only fully scientifically validated platform for the automatic extraction of digital biomarkers, capturing the multidimensional nature of verbal behavior, including the impact of automatic speech recognition (ASR) on downstream tasks and precise high-resolution measurements of expert-engineered indicators.
In-House Developed Tools
Our powerful analytics tools are entirely developed in-house, giving us unmatched control, flexibility, and security. By eliminating reliance on closed-source third-party software, we deliver a fully customizable, transparent, and adaptable solution tailored to meet the unique demands of our platform.
How it works
Exaia's Human-Centric and Explainable Artificial Intelligence (HCXAI) Platform
High-Precision Measurements of Digital Biomarkers.
Confidence Scores for Detected Mental Health Conditions.
Model Explainations for insights into Decision-Making Processes.
partner with us
Join Us in Our Mission to Transform Healthcare with Trustworthy AI and Digital Biomarkers
We invite researchers, clinicians, and healthcare leaders to collaborate with us in advancing a shared vision: developing objective, scalable, and non-intrusive AI solutions that leverage digital biomarkers derived from spontaneously produced speech and language. Through trustworthy AI, we aim to drive early detection, precise diagnostics, and continuous monitoring for conditions such as Alzheimer's disease, dementia, depression, bipolar disorder, generalized anxiety disorder, ADHD, and autism.
Collaborative Opportunities with Exaia's Digital Biomarkers
Early Detection & Risk Assessment
Identify patients with early susceptibility to neurodegenerative and psychiatric conditions, facilitating timely intervention.
Precision Diagnostics & Comorbidity Management
Harness biomarkers to accurately differentiate diagnoses in cases with overlapping symptoms, facilitating effective management of comorbid conditions and enhancing overall diagnostic accuracy.
Monitoring Disease Progression & Treatment Efficacy
Continuously assess disease progression and treatment impact over time, ensuring adaptive care and long-term management.
Testimonials
Dr. Ali Riza Günes
Senior Physician Center for Neurostimulation AJK-Neuss