Dr Maria Kyrgiou
Imperial College London
Awarded: £149,998
The challenge
Gynaecological sarcomas, commonly found in the uterus, make up about 13% of all sarcomas. Sadly, they often have a poor prognosis. Getting the right diagnosis early is fundamental to improving survival rates but current testing methods aren’t as accurate as they could be. Treatment with surgery can also be difficult, as it can be hard to tell the difference between normal tissue and the sarcoma with the naked eye. This can sometimes require a second procedure if not all is removed.
How will this project tackle this challenge?
This project aims to develop a piece of technology to help spot gynaecological sarcomas, to improve both diagnosis and surgery.
Imperial College London previously developed the iKnife for other gynaecological cancers. The tool works by analysing the smoke produced from surgical diathermy, a technique to cut and seal tissue such as blood vessels with a very hot instrument. The smoke is passed through an instrument called a mass spectrometer which ‘reads’ the chemicals to tell the difference between normal and cancer tissue. This method works but the burning damages tissue, which means that not all types of cells can be detected.
To analyse the cells without causing this damage, Professor Kyrgiou’s team aims to use a rapidly heating robot-guided laser (laser-REIMS) to remove a single layer of cells from the uterine wall. The laser can scan a large area, creating a ‘map’ of the normal and cancer cells to help surgeons before operating. Machine learning will also train the system to spot different subtypes of sarcoma.
What this means for people affected by sarcoma
The tool should help surgeons remove all of a tumour during surgery, meaning patients are less likely to have multiple operations. Longer term, the team hopes that the technology will help diagnose uterine sarcomas faster and earlier, and therefore improve survival rates from this group of rare cancers.
The laser can scan a large area, creating a ‘map’ of the normal and cancer cells to help surgeons before operating. Machine learning will also train the system to spot different subtypes of sarcoma.
