13.09.18

Progress for AI which could help in radiotherapy treatment

An artificial intelligence system which could help radiotherapy treatments for head and neck patients has shown promising early results.

Clinicians and researchers from DeepMind Health and University College London Hospitals (UCLH) are working on the system which, based off the initial testing, can perform to a similar standard as radiographers and oncologists but in a fraction of the time.

Tested in a non-clinical environment, the algorithm being developed can rapidly separate a range of anatomy from a scan, which means patients in the future could begin radiotherapy far more quickly. 

The current process takes up to four hours to manually mark-up cancerous and healthy tissues on CT scans of head and neck cancer patients.

Completing this mark-up process, known as segmentation, accurately is vital in cancer treatment so that the highest dose of radiation is given to the tumour whilst avoiding the healthy tissue.

This is particularly difficult for head and neck cancer, such as a mouth cancer, because tumours are extremely close to healthy structures.

Therapeutic radiographer and UCLH’s head of cancer services Kevin Sullivan said: “It is very early days for this research but the findings so far are really exciting.

“Having an automated system which could reduce part of the radiotherapy planning process from hours to minutes is a potentially game-changing development.

“In the future, it could enable us to treat patients more quickly and effectively, with the ultimate goal of improving the outlook for people diagnosed with head and neck cancer.”

The algorithm was developed using over 80 scans of 500 former head and neck cancer patients of UCLH who consented to their anonymised data being used for research purposes.

It has been tested on publicly available scans from the US and the next phase of this work will be to test the system in real clinical practice.

Just last week the health and social care secretary Matt Hancock outlined his vision for a digital NHS, unveiling £200m to help digitalise NHS trusts as well as announcing trials for an NHS app.

Enjoying NHE? Subscribe here to receive our weekly news updates or click here to receive a copy of the magazine!

Comments

There are no comments. Why not be the first?

Add your comment