Has AI rushed into oncology?

MRI Scan Why cancer-spotting AI needs to be handled with care
The Verge, Jan 27, 2020

“These days, it might seem like algorithms are out-diagnosing doctors at every turn, identifying dangerous lesions and dodgy moles with the unerring consistency only a machine can muster. Just this month, Google generated a wave of headlines with a study showing that its AI systems can spot breast cancer in mammograms more accurately than doctors … But for many in health care, what studies like these demonstrate is not just the promise of AI, but also its potential threat … in some areas where tech companies are pushing medical AI, this technology could exacerbate existing problems.”

Grant Thornton’s View Satish Gattadahalli
Satish Gattadahalli
Digital Health & Informatics
AI has the power to transform healthcare. That’s why AI solutions have attracted attention.

But an AI algorithm is not a black box that generates truth. It is a tool – and like any tool, it must be used in the right context by skilled people who understand its limitations.

In the case of cancer diagnosis, the article points out that “there’s no gold standard for cancer diagnosis, particularly early cancer.” Cancer diagnosis is complex work, and the implementation of AI technology in this work gives rise to ethical concerns that healthcare facilities must recognize and address. Facilities must determine the “who, what, when, where and why” of applying AI-driven tools in this and other areas, with governance and processes to ensure that the outputs are accurately interpreted and weighed by people who understand what the tool does well, and what it does not.

Then, facilities need to stay ahead of AI-driven tools by adapting their processes as the tools learn and change. When applied effectively and with ethical guardrails, AI can be integrated into the clinical decision support workflow. AI tools should be intended to assist clinicians as they validate and make evidence-based clinical decisions to optimize outcomes.

Consider this: How do you ensure that you understand the best way to implement your AI tools? How do you ensure that you keep the right processes in place?