AI challenge to improve neuroblastoma surgeries

Neuroblastoma is a tumor of the sympathetic nervous system that occurs mainly in young children. Every year, 25 children in the Netherlands are diagnosed with this tumor. In most children, surgery is one of the steps in treatment. Removing a neuroblastoma tumor as completely as possible is often a complex operation. The tumor is usually located in the child's abdominal cavity, between several organs and may be intertwined with these organs. In addition, in many cases the tumor envelops important blood vessels such as the aorta, vena cava or renal vessels.

For successful surgery, it is important that a surgeon already knows as much as possible about the location of the tumor during preparation. Therefore, MRI scans are made before the operation. These 'flat' 2D scans do not always show the complexity of the location and the relationship to surrounding organs and blood vessels. New solutions, such as 3D prints or holograms could better prepare a surgeon for surgery. With shorter surgery time and better tumor removal as possible results.

From 2D to 3D

To convert a scan to a 3D image, the images must be drawn in manually, also known as segmentation. This process is currently still time-consuming and error-prone because of the manual work. This segmentation is the challenge within the SPPIN challenge that artificial intelligence specialists from around the world can participate in through MICCAI 2023. Myrthe Buser, PhD candidate in the Wijnen and van den Heuvel group and coordinator of the SPPIN challenge: 'It would be very nice if together with the participants in the challenge we find solutions to automate the segmentation process. That way, we can use artificial intelligence to further improve the use of 3D models for surgeons by making them available faster and more reliable.'

To help AI specialists, MRI images of children with neuroblastoma are shared anonymously. Buser: 'It's just a few images, and with that, the participants are immediately presented with a challenge that we as researchers often face as well. Childhood cancer is rare and so is research data. The MICCAI network brings together the best artificial intelligence specialists so I am confident that great solutions will be devised.'

Within the Máxima, several researchers have united in an AI network . Together they are looking at how AI can contribute appropriately to the research being done at the Máxima.

Holograms and 3D models

Currently, the surgeons at Máxima are already using holograms and 3D prints in preparation for surgery on kidney tumors and neuroblastomas. Prof. Dr. Marc Wijnen:  'As a surgeon, it's really useful to know what to expect during an operation. With innovative techniques such as holograms and 3D prints, we have experienced that this is possible. Thanks to the SPPIN challenge, we can improve the creation of these models even further so that we can apply this more often. This way we will go into an operation as optimally prepared as possible which is expected to result in a shorter duration of the operation with fewer complications. It is great that AI specialists from all over the world contribute in this way to our mission to cure every child with cancer with optimal quality of life.'