Risk Assessment for COVID-19: An Artificial Intelligence Approach
About
Assessing the risk of potential diseases has increasingly attracted attentions in the field of modern medicine especially with the recent development of artificial intelligence (AI). The outbreak of COVID-19 requires our healthcare providers to optimally allocate our resources and design personalized treatment to individual patients. Artificial intelligence techniques can play an important role in clinical decision making by assessing the risk of patients based on their pathological profiles including electronic records, medical imaging and Omics data (e.g., metabolics, lipidomics and proteomics). The aim of this project is developing a computational approach to assess the risk of disease, where COVID-19 is the disease of our main interest.
There are many challenges we will face when designing our method.
- The first challenge is how to handle multi-modal datasets and construct a model to integrate them together. Data collected from different resources would be of different dimensionalities. Representation learning to map heterogeneous datasets into the same latent space should be developed.
- The second challenge is how to make our prediction results or model interpretable. Most existing AI methods draw their conclusions without providing any explanations, purely following a black-box approach. However, in the medical field, doctors need to know the evidence to support the conclusion from AI, inspiring the studying in interpretable AI.
- The third challenge is how to continuously refine the model when seeing new data. We need to make the proposed model be adaptive to different case studies. This is an interdisciplinary research that modern medicine approach and artificial intelligence techniques will be jointly investigated. Researchers from computer science and medicine will work closely together.
Required skills
The ideal candidate is who has a machine learning background or strong computational skills.
Principal Investigator
Assistant Professor, Department of Computer Science
Co-Investigators
Cheung On Tak Endowed Chair of Chinese Medicine, Teaching and Research Division
Associate Vice-President (Chinese Medicine Development)
Tsang Shiu Tim Endowed Chair of Chinese Medicine Clinical Studies
Vice-President (Research & Development)
Professor, Department of Computer Science