Mesothelioma is an unusual type of lung cancer that is extremely difficult to measure and detect. Because of its complex shape it requires highly skilled doctors to identify the tumours and draw them by hand on scans. This leaves room for user error and takes away valuable time that doctors could spend treating patients. <\/p>\n\n\n\n
To help solve this problem, experts at the University of Glasgow and Canon Medical Research Europe have created a prototype artificial intelligence (AI) system, which can detect and measure Mesothelioma tumours on CT scans without any human input. <\/p>\n\n\n\n
Learn how experts trained the AI system to automatically detect tumours by using existing scans that were marked by doctors. Could this AI tool soon help doctors treat cancer patients with greater precision and for less money? Could it help improve clinical trials to find new drug treatments more quickly?<\/p>\n\n\n\n
AI system able to detect and measure complex tumours without any human input<\/strong> <\/h2>\n\n\n\n
Mesothelioma is a cancer of the lining of the lung caused by inhalation of asbestos, a building material commonly used in construction. Glasgow has the highest incidence of Mesothelioma in the world, due to the historical use of asbestos in its shipbuilding industry. <\/p>\n\n\n\n
Asbestos has been banned in the UK since 1999, however we expect to see Mesothelioma cases continue to rise as it can take 30-50 years for the disease to develop after asbestos exposure. <\/p>\n\n\n\n
Countries such as Brazil, Russia, India, and China still use asbestos in their construction industries, and so we expect to see global cases rise too. <\/p>\n\n\n\n
Mesothelioma is unusual because it doesn\u2019t grow like a round ball, like most tumours. Instead, it grows like a \u2018rind\u2019 around the surface of the lung, forming a complex shape. <\/p>\n\n\n\n
<\/figure><\/div>\n\n\n\n
Mesothelioma shown within yellow loop<\/em><\/p>\n\n\n\n
Think about the orange peel being the tumour rather than a seed in the centre of the orange. <\/p>\n\n\n\n<\/figure>\n\n\n\n
Mesothelioma shown in green<\/em> <\/p>\n\n\n\n
This makes it difficult to detect and measure Mesothelioma on scans and requires highly skilled doctors to identify and quantify the size of the tumour, and any changes with treatment. It involves drawing the tumour on the scans by hand based on their own experience and judgement. This is time consuming, leaves room for user error, and uses up time the doctor could spend treating patients. <\/p>\n\n\n\n
Experts at the University of Glasgow and Canon Medical Research Europe (a Scottish firm specialising in next generation medical imaging software) have created a prototype AI system able to automatically find and measure Mesothelioma on CT scans without any human input. <\/p>\n\n\n\n
The AI was trained by showing it over 100 CT scans, on which an expert clinician had drawn around all areas of the tumour which showed 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 on its own. <\/p>\n\n\n\n
<\/figure><\/div>\n\n\n\n
<\/figure><\/div>\n\n\n\n
Human drawing in red AI drawing in green<\/em><\/p>\n\n\n\n
The AI tool streamlines tumour measurements, potentially making clinical trials of new drugs more accurate, less time-consuming and therefore less expensive. <\/p>\n\n\n\n
This pilot project was funded by Cancer Innovation Challenge, and has been successful in leveraging a multi-million Cancer Research UK Accelerator Award for a Mesothelioma Research Network called PREDICT-Meso. <\/p>\n\n\n\n
PREDICT-Meso brings together expert mesothelioma researchers from across Europe to design and validate new therapies and diagnosis techniques for Mesothelioma. <\/p>\n\n\n\n
After further work to test and validate the AI tool as part of PREDICT-Meso, the AI technology may soon be available to help doctors measure Mesothelioma on scans during treatment with greater precision and at a reduced cost. <\/p>\n\n\n\n
The successful results of the project will provide a strong foundation for similar tools to be developed in the assessment of other cancers. <\/p>\n\n\n\n
![\"\"](\"https:\/\/curiousabout.glasgowsciencecentre.org\/innovation\/wp-content\/uploads\/sites\/3\/2021\/05\/PDFImage1-Mesothelioma_Loop-300x272.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/curiousabout.glasgowsciencecentre.org\/innovation\/wp-content\/uploads\/sites\/3\/2021\/05\/PDFImage2-TumourVolumebyHuman-1024x546.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/curiousabout.glasgowsciencecentre.org\/innovation\/wp-content\/uploads\/sites\/3\/2021\/05\/PDFImage3-HumanvsAI.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/curiousabout.glasgowsciencecentre.org\/innovation\/wp-content\/uploads\/sites\/3\/2021\/05\/PDFImage4-HumanvsAI.png\")
<\/figure><\/div>\n\n\n\n