Focus Areas 

The UBC AI and Health Network is currently focused on research that leverages the latest advances in natural language processing (NLP), large language models (LLMs) and foundation models (FMs) to address real-world challenges in two key pillars: 1) clinical practice and patient experience and 2) biomedical innovation and translational sciences. As the Network grows, we expect to add additional pillars to expand our scope in a focused and strategic way. 

Three cross-cutting themes underpin the Network’s interdisciplinary research in AI and health:   

  1. Data-sharing and privacy practices;  
  2. End-user acceptance of AI in healthcare; and  
  3. Province-wide education, training and upskilling in both the STEM (science, technology, engineering, and mathematics) and clinical/health fields.  

Pillar 1: Clinical Practice and Patient Experience 

Clinical documents such as patient charts, pathology reports and radiology reports contain a wealth of valuable information; however, these documents are in natural language format that requires human experts to read and interpret them. Similarly valuable information can be found in user-generated content, which refers to all the speech and text information provided by patients, their families and caregivers on their needs, feelings, moods and preferences.  

Research aligned to this pillar involves developing and using NLP tools to automatically extract key pieces of information from clinical documents to expedite triage or to analyze user-generated data for early detection and treatment of many medical conditions, such as cancer, mental illness, neurodegenerative diseases and neurological development delays. 

Pillar 2: Biomedical Innovation and Translational Sciences 

Research aligned to this pillar involves developing and using NLP tools to expedite the discovery, development and adoption of new therapeutics. The same technology that builds LLMs for English sentences can be used to build foundation models for biological structures such as DNA, proteins and metabolites to advance protein and drug design. These AI tools can be widely used in protein and biochemistry analysis to make the drug development process faster and more cost-efficient. In addition, NLP tools can be used to expedite the regulatory processes involved in bringing new therapeutics into the healthcare system. Developing these tools to support researchers’ understanding of the pathway to regulatory approval will highlight key challenges and opportunities for the efficient and timely translation of medical innovations into clinical practice and inform deliberations on value, safety and cost-effectiveness.  

Theme 1: Data Sharing and Privacy 

Leveraging AI tools for health system transformation requires massive amounts of data. Individual researchers often need to share data with their collaborators to augment their sample sizes and diversity. Similarly, health service providers see immense potential in giving researchers access to their data to support research and innovation. Both stakeholder groups recognize the critical importance of preserving patient privacy. Research aligned with this theme will develop privacy-preserving approaches and tools for health data sharing, as well as tools for synthetic data generation, which offers the additional advantage of data augmentation. The latter will prioritize representation of equity-deserving communities in training data and AI tools to mitigate biases and address the needs of the diverse patient populations in B.C. 

Theme 2: End-User Acceptance of AI in Healthcare 

Facilitating an open dialogue about the ethical, equitable, effective and economical use of AI in patient care and biomedical research is critical. The Network will ensure innovative approaches to early detection, triage, disease management and treatment are acceptable by end users and adopted into the health system. Research aligned with this theme includes interdisciplinary approaches to the ethical dimensions of using emerging technologies and AI in healthcare delivery, research, clinical trial design and supportive decision-making in medicine, as well as implementation science—the methods and strategies to incorporate findings into clinical practice, research and health system planning. 

Theme 3: Province-wide Education, Training and Upskilling 

UBC’s model of distributed medical/health professional education and continuing professional development reaches students and healthcare providers in communities across the province. Activities aligned with this theme will leverage this framework for growing understanding and acceptance of AI in health care. This work will be informed and strengthened by deep and trusting relationships with clinicians and health system partners across the province to support implementation science.