This Science News Wire page contains a press release issued by an organization and is provided to you "as is" with little or no review from Science X staff.

Robots learn to help doctors spot breast cancer

March 15th, 2018 Hayley Jarvis

Scientists are using artificial intelligence to build an ultra-accurate MRI scan-understanding system that could change the way doctors diagnose breast cancer.

In a two-year experiment, doctors at South Tees Hospitals NHS Foundation Trust will use a cutting-edge technology to analyse thousands of breast scans from all the MRI machines in the UK.

IntelliScan, which will be developed by Brunel University London, First Option Software and Teesside University, combines deep machine learning and advanced image-processing smart algorithms.

It will automatically flag up key watch points on each potential case, which will let radiologists handle many more images faster and more accurately, and help save lives by detecting tumours earlier.

"The technique will provide unique information to support doctors during the diagnostic process," said Brunel's Sergio Malo. "And the machine learning algorithm will use the doctors' feedback for further training and for continuous improvement of the detection tool across the UK."

Once fully developed, the plan is to link the system to breast scans from all MRI systems UK-wide, so it can detect abnormalities and categorise them by severity.

Its algorithms will also help doctors predict how well treatment will work on individual patients by comparing their consecutive scans.

IntelliScan's digital image processing and artificial intelligence decision system will be developed at Brunel Innovation Centre, which specialises in developing algorithms for pattern, defect and anomalies detection, through modelling machine learning and image processing. It has developed expertise in similar technologies such as behaviour modelling and advance inspection of structures.

Missed or late diagnosis is behind 20% of cancer deaths in the UK, which has the some of the worst cancer survival rates in Western Europe. The £830,000 two-year project, funded by Innovate UK, will completely change how the NHS uses MRI scans to diagnose and treat breast cancer. It combines specialist image processing and machine learning with the Internet of Things, through cloud hosting and remote data processing.

Dr. Jianxin Gao, Director of Teesside University's Healthcare Innovation Centre, added: "This is an exciting project and one that could make considerable and positive changes to the healthcare sector – improving the accuracy and speed of breast cancer MRI analysis. Early detection is crucial to survival and this technology could help to save the lives of thousands of patients."

Hospital radiographers will automatically get highly digitally advanced reports, allowing them to help doctors dramatically improve outcomes for patients.

"The system integrates a series of visualisation, data processing, data communication and decision-support systems which will enable it to dramatically improve access to breast healthcare and cancer treatment compliance," said Professor Tat-Hean Gan, Brunel Innovation Centre's director.

More information:
The research team is: Brunel Innovation Centre at Brunel University London, First Option Software (SME, project coordinators), Healthcare Innovation Centre (HIC) at Teesside University and South Tees Hospitals NHS Foundation Trust.

IntelliScan has the potential to generate £18.5 million in revenues and £15.4 million in profits by 2028 and create 49 direct and indirect jobs related to the outcome of the project.

Provided by Brunel University

Citation: Robots learn to help doctors spot breast cancer (2018, March 15) retrieved 8 September 2024 from https://sciencex.com/wire-news/282562098/robots-learn-to-help-doctors-spot-breast-cancer.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.