Therapy planning device

A training system for training a machine learning module, the training system comprising:

• a therapy application parameter providing unit configured to provide at least one therapy application parameter defining an application of a thermal ablation therapy to tissue of a subject,
• a therapy result distribution providing unit configured to provide at least one therapy result distribution defining at least one distribution of at least one parameter being indicative of at least one obtained effect of the thermal ablation therapy which has been applied in accordance with the at least one therapy application parameter defining the applied thermal ablation therapy,
• a tissue parameter distribution providing unit configured to provide at least one tissue parameter distribution which has been generated based on an magnetic resonance scan of the tissue of the subject, before the thermal ablation therapy has been applied to the tissue,
• a training unit configured to train the machine learning module by using as training input at least the at least one provided tissue parameter distribution and the at least one provided therapy result distribution and as training output the at least one provided therapy application parameter.

A therapy planning method for planning a thermal ablation therapy to be applied to tissue of a subject, the therapy planning method comprising:

• providing at least one tissue parameter distribution, which has been generated based on an magnetic resonance scan of the tissue, by a tissue parameter distribution providing unit,
• providing at least one therapy goal distribution defining a distribution of at least one parameter being indicative of a desired effect of the therapy by a therapy goal distribution providing unit,
• providing a machine learning module which has been trained to output a therapy application parameter defining the application of the therapy based on at least one input tissue parameter distribution and at least one input therapy goal distribution, and
• planning the therapy by determining the at least one therapy application parameter based on the provided at least one tissue parameter distribution, the provided at least one therapy goal distribution and the trained machine learning module by a therapy planning unit.

A training method for training a machine learning module, the training method comprising:

• providing at least one therapy application parameter defining an application of a thermal ablation therapy to tissue of a subject by a therapy application parameter providing unit,
• providing at least one therapy result distribution defining at least one distribution of at least one parameter being indicative of at least one obtained effect of the thermal ablation therapy which has been applied in accordance with the at least one therapy application parameter defining the applied thermal ablation therapy by a therapy result distribution providing unit,
• providing at least one tissue parameter distribution which has been generated based on an magnetic resonance scan of the tissue of the subject, before the thermal ablation therapy has been applied to the tissue, by a tissue parameter distribution providing unit,
• training the machine learning module by using as training input at least the at least one provided tissue parameter distribution and the at least one provided therapy result distribution and as training output the at least one provided therapy application parameter by a training unit.