Meeting MI - Radiologists Behind AI: Dr. Farzad Khalvati, Chair in Medical Imaging and Artificial Intelligence, SickKids
Meeting MI – Driving AI offers the Department of Medical Imaging the opportunity to meet the members of MI and learn about their work with artificial intelligence (AI).
Name: Farzad Khalvati
Division: Non Clinical
Academic Rank: Associate Professor
Hospital site: Senior Scientist, Chair in Medical Imaging and Artificial Intelligence
1. What role do you play in the Department’s work advancing AI research and integration into radiology?
I am the director of Intelligent Medical Informatics Computing Systems Lab (IMICS.ca) at SickKids where I collaborate closely with radiologists to develop, validate, and deploy AI tools for early diagnosis and prognosis, and individualized treatment planning. Using multi-modal data such as medical imaging and health informatics, our goal is to advance precision medicine and improve patients care. My research also focuses on creating interpretable, human-centered AI for medical imaging and ensuring equitable access to AI-driven healthcare solutions.
2. What are some of the highlights and challenges you’ve experienced working with AI?
AI performance has improved significantly, but making it explainable and trustworthy remains a challenge. Explainable AI helps radiologists understand how decisions are made, while trustworthy AI provides confidence measures, allowing them to determine when it is safe to rely on AI. Another key challenge is seamlessly integrating AI into existing workflows with minimal disruption.
3. Looking ahead, what are you most excited to see with the integration of AI into medical imaging?
Integrating AI into medical imaging presents numerous opportunities to enhance healthcare delivery. It can help bridge the gap in access to high-quality care, particularly in regions with a shortage of specialized radiologists, thereby promoting more equitable healthcare. By automating routine and repetitive tasks, AI allows radiologists to focus on more complex and high-value aspects of patient care, improving efficiency and reducing burnout. Additionally, AI can augment radiologists’ capabilities, enabling them to perform advanced tasks such as imaging-based molecular diagnosis—an area that may currently be beyond human interpretation. Ultimately, AI has the potential to transform medical imaging by making it more precise, efficient, and widely accessible.
4. Tell us something about you that might surprise your colleagues!
I have long been passionate about literature, particularly ancient Persian texts and I have co-authored a book “The Hidden Teachings of Rumi”. This interest has led to an exciting collaboration with the Department of Near and Middle Eastern Civilizations at the University of Toronto, where we are exploring how AI can be applied to Persian literature and broadly to humanities field. As we enter a new era in digital humanities, AI is opening up unprecedented opportunities for researchers to analyze, interpret, and preserve historical texts in ways that were previously unimaginable. From uncovering hidden patterns in classical works to making vast collections more accessible, AI is transforming the study of humanities, bridging the past with the future.