The initial dataset consists of more than 15,000 data already held by the lead partner, validated through scientific publications and correlation statistics, on cognitive and biomolecular phenotype derived from multiple repeated tests on animal models of cognitive disorder.

An initial model will be built on this data, which will subsequently be implemented with data derived from cognitive profiling, plasma biochemistry, and metagenomic analyses derived from the gut microbiota in animal models of chemobrain.
The final product will be a prototype demonstrator (software) using the ML/AI predictive model developed and validated above.

The ultimate goal is thus to identify patients at risk of chemobrain from data collected by non-invasive or minimally invasive approaches at the beginning of treatments, in order to develop personalised preventive interventions.