Better analysis of scans to adapt radiotherapy
Thierry Colin - © Inria / photo H. Raguet
The MC2 team at Inria-Bordeaux Sud-Ouest has developed an adaptive radiotherapy technique for automatic monitoring of organs using low-resolution control scans. The benefit for the patient is that it avoids overdoses in at-risk organs.
In the treatment of certain tumours, radiotherapy is all planned in one go, with no adjustment as the sessions progress. Yet patient weight loss and the movement of vital organs over time may increase the dose actually received by the patient so that it exceeds the planned dose. A overdose to at-risk organs such as the duodenum may have terrible consequences.
A long-term collaborator of Guy Kantor's team in the Radiotherapy department of the Institut Bergonié in Bordeaux, Thierry Colin 's MC2 project team (Modelling, control and computations for fluid mechanics and biology), turned its attention to adaptive radiotherapy for retroperitoneal sarcomas.
Thanks to image processing, a first stage involving around ten patients showed that the movement of organs such as the kidneys was greater than doctors thought and that their monitoring indicators (such as weight) were not pertinent. The study has been presented at several radiotherapy conferences, with the next presentation planned for the ASTRO conference in San Francisco in September 2014.
MC2's mathematical models could also impact medical imaging equipment, by offering a better vision of the illness through more uniform, standardised readings.
The next stage is the development of techniques for the automatic monitoring of at-risk organs using low-resolution control scans. Contact has also been made with a company that is developing dosimetry software in order to test the re-planning of radiotherapy.
Finally, two new research avenues are to be launched on primitive lung and kidney tumours, with the same objectives regarding the monitoring and assessment of the patient's response to treatment. The MC2 team's ambition is to use new medical imaging techniques to garner more information about the tumour and thus improve the precision of its models for each specific pathology.