A new artificial intelligence (AI) model can predict a cancer patient's survival with more accuracy than in the past. The findings were published Thursday in JAMA Network Open.
Researchers from the University of British Columbia and BC Cancer developed the model using natural language processing (NLP) — technology that allows AI to understand complex human language — to analyze an oncologist's notes and pick up on unique clues during consultation visits.
“The AI essentially reads the consultation document similar to how a human would read it,” said lead author Dr. John-Jose Nunez, a psychiatrist and clinical research fellow with the UBC Mood Disorders Centre and BC Cancer, in a press release.
“These documents have many details like the age of the patient, the type of cancer, underlying health conditions, past substance use, and family histories. The AI brings all of this together to paint a more complete picture of patient outcomes.”
Based on research findings, the model predicted survival rates with more than 80 per cent accuracy.
In the past, cancer survival rates were calculated retrospectively, and categorized by few generic factors like cancer site and tissue type. And unlike previous calculations, the new AI model is applicable to all cancers.
“Predicting cancer survival is an important factor that can be used to improve cancer care,” said Nunez.
“It might suggest health providers make an earlier referral to support services or offer a more aggressive treatment option upfront. Our hope is that a tool like this could be used to personalize and optimize the care a patient receives right away, giving them the best outcome possible.”
More importantly, the model was trained on B.C. data, which means it can serve as a tool for B.C. patients and their cancer survival rates.
Because NLP models are portable and don't require structured data sets, they have the possibility of being applied in cancer clinics across Canada and the world, explained Nunez.
“As medicine gets more and more advanced, having AI to help sort through and make sense of all the data will help inform physician decisions. Ultimately, this will help improve quality of life and outcomes for patients.”