An Image of The Future

18 March 2022

Artificial intelligence could help us to assess scans of patients with bone disease, thanks to an innovative software toolkit developed at the BRC.

Prof Koh
Pictured: Professor Dow-Mu Koh, Consultant Radiologist at The Royal Marsden

Bone disease has a complex relationship with cancer. Some cancers, such as breast and prostate, can spread to the bones, which is known as metastatic bone disease. Also, bone disease is a complication of other cancers, including multiple myeloma.

Doctors typically keep track of bone disease and how it responds to treatment using CT and bone scans. However, these can be less reliable than some more complex techniques, such as whole-body MRI scanning (WB-MRI). WB-MRI scans can produce more than 1,000 images, which are extremely time-consuming for a radiologist to read and report.

To improve the efficiency of WB-MRI, Professor Dow-Mu Koh, Consultant Radiologist at The Royal Marsden, and Dr Matthew Blackledge, Team Leader of the Computational Imaging Team at The Institute of Cancer Research, London, have developed a software toolkit that automatically processes these images. Developed in partnership with Mint Medical, the toolkit uses artificial intelligence (AI) to detect regions of disease in MRI scans by tracking changes in the random motion of water molecules in bone tissue, which can differentiate between healthy and abnormal areas.

Through this technique, the software rapidly highlights regions of disease in the images and produces a detailed report. It has been tested in an early trial and has shown to be reliable and accurate in assessing the extent of bone disease in prostate cancer patients. Dr Blackledge says: “Each body responds differently to cancer treatment. If we can measure specific regions of disease that are not responding to treatment earlier, then this gives us a window of opportunity to do something.”

In early 2021, the team was awarded almost £2 million in funding from the National Institute for Health Research (NIHR) Invention for Innovation (i4i) Challenge programme to test the software in two large multi-centre clinical trials, involving patients with bone disease from prostate cancer and myeloma. Professor Koh says: “It is important that we continue to improve and develop new ways of tracking and treating cancer. We are grateful for the i4i funding so that we can continue to investigate our toolkit.”

This project forms part of The Royal Marsden’s AI Imaging Hub, which investigates digital solutions to enhance our understanding of cancer and improve patient outcomes