Multimodal Disease Cohort

Integrating diverse data sources for advanced disease modeling and insights into gene-drug-phenotype relationships.

Innovative Disease Cohort Solutions

We specialize in constructing disease cohorts using multimodal longitudinal data to enhance understanding of complex diseases through advanced analytics and modeling.

A group of medical professionals and possibly officials are gathered around a patient in a hospital ward. The patient is seated on a hospital bed, engaging with the people around him. Several men in uniform and plain clothes are present, suggesting a formal visit or inspection.
A group of medical professionals and possibly officials are gathered around a patient in a hospital ward. The patient is seated on a hospital bed, engaging with the people around him. Several men in uniform and plain clothes are present, suggesting a formal visit or inspection.
A medical monitor with a digital display shows various patient statistics, including heart rate and other vital signs. The device is placed above a printed graph output, likely representing a heart rate or other medical measurement over time. The screen features bright numeric and graphical displays.
A medical monitor with a digital display shows various patient statistics, including heart rate and other vital signs. The device is placed above a printed graph output, likely representing a heart rate or other medical measurement over time. The screen features bright numeric and graphical displays.
The image displays a digital screen with COVID-19 statistics, including the numbers of recovered, deaths, and suspected cases. A map with blue circular markers is visible on the right side, indicating different locations.
The image displays a digital screen with COVID-19 statistics, including the numbers of recovered, deaths, and suspected cases. A map with blue circular markers is visible on the right side, indicating different locations.

Our Approach to Research

Utilizing spatiotemporal graph neural networks, we model dynamic interactions among biomarkers to uncover insights into disease mechanisms and treatment pathways.

Multimodal Data Solutions

Construct disease cohorts using advanced multimodal longitudinal data for comprehensive analysis and insights.

A close-up view of a digital medical monitor displaying vital signs including a green electrocardiogram (ECG) line, a yellow respiratory rate line, and numerical values with a red background indicating a concerning level.
A close-up view of a digital medical monitor displaying vital signs including a green electrocardiogram (ECG) line, a yellow respiratory rate line, and numerical values with a red background indicating a concerning level.
A medical monitor displays various vital signs and readings, including an ECG line showing heart activity, numbers indicating SPO2 levels, and blood pressure. The screen is prominently labeled with 'DEMO', and the monitor is manufactured by Mindray.
A medical monitor displays various vital signs and readings, including an ECG line showing heart activity, numbers indicating SPO2 levels, and blood pressure. The screen is prominently labeled with 'DEMO', and the monitor is manufactured by Mindray.
Dynamic Network Modeling

Utilize spatiotemporal graph neural networks to model interactions and optimize cross-modal data alignment.

Knowledge Graph Development

Leverage GPT-4 for relation extraction to create a gene-drug-phenotype knowledge graph from PubMed literature.

Insights

Leveraging multimodal data for disease understanding.

© 2025. All rights reserved.