Assistive AI in cancer care - DHW speaker - Health Informatics New Zealand

eHealthNews.nz: AI & Analytics

Assistive AI in cancer care - DHW speaker

Monday, 28 August 2023

Share |

NEWS - eHealthNews.nz editor Rebecca McBeth

Medical AI is hugely helpful to support doctors with repetitive tasks, but will not take clinicians’ jobs any time soon, says Raj Jena, oncologist at Cambridge University Hospitals NHS Foundation Trust.

Jena is a keynote speaker at Digital Health Week NZ 2023. He led the development of a new AI tool, OSAIRIS, which significantly cuts the amount of time a doctor needs to spend drawing around healthy organs on scans by calculating where to direct therapeutic radiation beams.

The technology is being used at Addenbrooke’s Hospital for prostate and head and neck cancers, where it is reducing radiotherapy wait times, and is the first cloud-based AI technology to be developed and deployed within the NHS.

You’ve read this article for free, but good journalism takes time and resource to produce. Please consider supporting eHealthNews by becoming a member of HiNZ, for just $17 a month.

Jena believes there will be increased use of assistive AI in cancer care in the form of clinical documentation, safety checks and patient dashboards, but says we are a long way from generalisable AI that could replace the insights of a clinician.

He says medical AI is still very primitive and task specific, meaning it can only reproduce the insights that were provided when it was trained for a specific task.

“Even large language models like ChatGPT, which on the surface seem to generate intelligent answers, have limited applications in medicine,” he tells eHealthNews.

“The more interesting application for the future will be in the use of AI to integrate data from multiple sources, particularly genomic information.

“This form of AI will be used to make predictions of outcome and therapy response, but is still at an early stage.”

New Zealand’s small population means it can be agile in its adoption of AI and cloud computing enables this, says Jena.

“Cloud computing avoids hospitals having to spend millions of pounds on expensive computers and means you can setup a new AI system in a hospital quickly and cheaply,” he says.

“Agility also means that standardisation is easier, as it would be easier for healthcare providers to agree on common solutions or AI implementation for everyone’s benefit.”

His vision for three to five years’ time, is to be able to preview an AI generated radiotherapy treatment plan with his patient in clinic, to help them better understand how the radiotherapy might affect them.

“I expect AI to provide more automation, and I expect this to translate to more time in front of the patient,” he says.

OSAIRIS was developed in collaboration with Microsoft Research.

Hear more insights from Raj Jena at Digital Health Week 2023 this November 27-30 in Hamilton.