September 19, 2019
Learn how a new technology like artificial intelligence (AI) has the potential to improve the delivery of cancer care, now and in the future
As the number of Canadians diagnosed with cancer increases, new models of delivering care are needed that improve access and efficiency and reduce costs. The Partnership is the steward of the recently refreshed 2019 to 2029 Canadian Strategy for Cancer Control (the Strategy) , which states that technology needs to play a more prominent role in improving cancer outcomes for Canadians.
AI tools have the potential to make cancer care more efficient and less labour-intensive, reducing costs to the system
The following key areas are where AI is being used or explored for cancer care:
- Analyzing data in support of detecting cancer earlier and, or identifying people at higher risk of cancer. Some of these approaches use less-invasive methods than traditional imaging and biopsies.
- Supporting cancer diagnosis, potentially making it more efficient and enhancing access by reviewing diagnostic and clinical images, segmenting images, highlighting suspicious regions shown in images for review and, or classifying findings as benign or malignant. Also, pathology findings can be classified and biomarkers identified as associated with imaging features (radiomics).
- Supporting the planning and decision-making for cancer treatment by assembling and reviewing patient clinical data, published literature and, or other medical evidence to inform individual treatment; and adapting a personalized approach to cancer treatment and follow-up care by predicting disease progression, survival and treatment response.
- Better identifying and proactively managing symptoms and complications cancer patients may experience.
- Making the care process more efficient by automating tasks that were previously done by humans, such as radiotherapy planning and scheduling for health care providers and patients by capturing hands-free documentation from providers in real time using AI-powered natural-language processing.
- Supporting quality improvement by extracting data and real-world evidence from electronic health records (EHRs) to inform quality indicators and monitoring, and drive treatment decisions and system change.
- Improving the patient experience by more easily giving them information and tailoring support to their specific needs.
Each of those key areas are explored in more detail in the report with explanations of how AI technologies and algorithms are being used or piloted in practice.
AI innovation examples currently in use or testing
The report also includes examples of specific AI innovations that are being used or tested in real-world settings. Such examples help to illustrate the possibilities for using AI to improve outcomes for cancer patients:
- Doctor AIzimov
- Watson for Oncology
- Watson for Clinical Trial Matching
- Precise MD
AI implementation considerations
This environmental scan also identifies many important elements that need to be considered and addressed to support widespread implementation of AI technologies:
- Ethical issues
- Access to data, patient privacy and data security
- Transparency, replication, validation, and testing
- Impact on the health workforce
- Costs and cost-effectiveness