TBValid collection: Pulmonary tuberculosis validation collection

The TBValid dataset comprises 870 digital patients with different profiles, each with fixed Age, BMI and MtbSputum. Each is identified by a vector of features involving biological and pathophysiological parameters to roughly represent different profiles in the population and initial bacterial load. Individual patient data collected during a clinical trial have been transformed into aggregated data, which are already irreversibly anonymised. Subsequently, these aggregated data have been sampled through the procedure described in “Generation of digital patients for the simulation of tuberculosis with UISS-TB”, doi: 10.1186/s12859-020-03776-z. The obtained derivative dataset, owned by its creators, does not constitute sensitive data according to European laws, and it is impossible with this dataset to re-establish the identity of the patients enrolled in the original clinical trial.