| Title: | Multi-modal AI representation learning to accelerate discovery of composite biomarkers |
| Lead Institution: | IBM Research |
| Principal investigator: |
Application 95318
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
About
With the increasing volume and variety of health data that is being collected, there is an important need for tools to help make sense of the combined data and use it to improve patient health. This three-year project will leverage the large, rich, and diverse health data in the UK Biobank and aim to 1) develop machine learning tools that can efficiently combine large volumes of data from a variety of data types into robust and comprehensive representations of patient health state, and 2) demonstrate the utility of the representations by using them to assess current status and predict future risk at the individual patient level for multiple diseases. The discovered insights can enable earlier detection of disease, improve clinicians' ability to identify people who would benefit from earlier interventions or participation in appropriate clinical trials, and assess the impact of specific therapies for specific patients. The developed computational tools will be made available to the scientific community to help accelerate scientific discovery in additional clinical areas.2 Publications
| Pub ID | Title | Author(s) | Year | Journal |
|---|---|---|---|---|
| 9962 | Fusion of biomedical imaging studies for increased sample size and diversity: a case study of brain MRI | Matias Aiskovich (+8) | 2024 | Frontiers in Radiology |
| 8960 | Subtyping Gastrointestinal Surgical Outcomes from Real World Data: A Comprehensive Analysis of UK Biobank. | Uri Kartoun (+11) | 2024 | AMIA Annual Symposium Proceedings |
Enabling scientific discoveries that improve human health