Problem 3

We propose the development of an algorithm for automatic segmentation and guidance of cryoablation intervention for desmoid tumors. The process involves administering an extremely cold freezing agent (liquid nitrogen, liquid nitrous oxide or compressed argon gas) through a fine needle system called a cryoprobe that freezes and destroys the abnormal tissue.

During a cryoablation procedure, one or more cryoprobes are inserted into the affected area at different points in the tumor. Once the needle is inserted, the freezing agent circulates inside the cryoprobes, creating an ice ball around the target tissue. The ice balls cover the entire tumor, as well as a margin of healthy tissue, to ensure that the area has been completely treated.


The optimal temperature for cryoablation is around -40 degrees Celsius. The frozen spheres have different diameters and the temperature expands radially, from low to high. Therefore, given a point inside the tumor, the temperature at this point would be given by the influence/diffusion of the different spheres in their field of action.


Based on an automatic segmentation of the tumors, the proposed problem is the optimal selection of the minimum number and position of cryoablation balloons to give a uniform temperature throughout the tumor volume.