Site navigation

New AI Tool Can Track Asbestos Cancer

David Paul

,

Asbestos Cancer

The AI tool used to assess patients receiving mesothelioma treatment could “revolutionise” care.

The Data Lab has today released details of a new project using AI to assess and treat mesothelioma, or ‘asbestos cancer’.

Scottish medical imaging software firm Canon Medical Research Europe and the University of Glasgow are set to publish clinical findings from a study evaluating a new cancer assessment tool, developed as part of the Cancer Innovation Challenge.

A study team have created a prototype that can automatically find and measure mesothelioma on CT scans. These scans are then used by the trained AI to assess patient’s response to drug treatments like chemotherapy.

The AI was trained by showing it over 100 CT scans, on which an expert clinician had drawn around all areas of the tumour – showing the AI what to look for. The trained AI was then shown a new set of scans and was able to find and measure the tumour extremely accurately, without any human input.

Commenting on the AI tool, Keith Goatman, Principal Scientist at Canon Medical, said: “The speed and accuracy of the AI algorithm could have a wide-reaching impact on mesothelioma treatment.

“Accurate tumour volume measurements are much too time-consuming to perform by hand. Automating these measurements will open the way for clinical trials of new treatments, by detecting even small changes in the tumour size. Ultimately, it could be used routinely in hospitals to decide the best treatment for each individual.

“The funding and support from the Cancer Innovation Challenge has been vital in bringing this idea to life, and we are looking forward to continuing our work with the excellent team at the University of Glasgow in the years to come.

“This work is a strong first step towards real change in the treatment of all cancers – not just mesothelioma.”

Scotland currently has the highest incidence of mesothelioma in the world, a reflection of the historical use of asbestos in many UK industries, including shipbuilding and construction.

At present, treatment options for mesothelioma are limited and clinical trials are critical for the discovery of new, more effective treatments. According to researchers, the AI tool “streamlines” tumour measurements, potentially making clinical trials of new drugs less expensive, less time-consuming and more accurate.


Recommended


Professor Kevin Blyth, Professor of Respiratory Medicine at the University of Glasgow, said: “To our knowledge, this study is world-leading in its successful use of AI to assess treatment response in mesothelioma.

“Using external data sets to validate our results, we have shown that tumours can be accurately measured by AI, giving us a new tool that will help us make better decisions for patients on treatment and reducing barriers to the development of new treatments in clinical trials.

“The results, which are testament to the expertise of Canon Medical and made possible by the Cancer Innovation Challenge funding, have acted as a springboard towards our next project, the PREDICT-Meso Accelerator, which is now allowing us to further develop the AI so that it can start benefiting patients soon.”

Steph Wright, Director of Health & Wellbeing Engagement at The Data Lab, added: “The work to develop this world-leading tool from Canon Medical and the University of Glasgow, represents an incredibly exciting healthcare innovation.

“Not only does it have the potential to revolutionise mesothelioma cancer care through more targeted treatment, but it may also be able to be applied to a number of other cancer types in the future.

“It’s been a privilege to play a part in helping to deliver the Scottish Funding Council’s Cancer Innovation Challenge initiative, supporting and spotlighting the companies carrying out valuable work that can help make Scotland a leader in data-driven cancer support. Through projects like this, we really can show that data saves lives.”

David Paul

Staff Writer, DIGIT

Latest News

%d bloggers like this: