The Virtual Child

virtual child

The Challenge

Solid tumours in children

The first ever learning and data-driven computer model of normal and cancerous nervous system development, to support biological and therapeutic discovery

Team Leads

Stefan Pfister DKFZ, Germany Richard Gilbertson University of Cambridge, UK

Countries involved

Germany, UK, US, Canada, Netherlands

The Idea

A number of barriers hinder the development of treatments for childhood tumours of the central and peripheral nervous system: their rarity makes them difficult to study, we need to understand more of their unique biology, and observations in adults do not translate to children.

This innovative team hopes to remove these barriers by building The Virtual Child – the first data-driven computer model of normal and cancerous nervous system development. Open-source and incorporating massive scale multimodal data, this ever-learning model could unlock a deeper understanding of childhood cancer and help to identify and prioritise new treatment avenues by utilising the concept of virtual clinical trials.

Drawing on a unique combination of world-leading skills – including childhood oncology, neural development, evolution and AI – the Virtual Child team brings fresh thinking to childhood cancer research and could transform the way we research these challenging diseases. 

 Having studied children’s brain tumours for 30 years, we never dreamed we’d be in with a chance to totally change the field.