A new approach to cancer treatment using belief-space control and active inference shows promise in personalizing therapies while managing measurement budgets. This method, tested on real clinical data, demonstrates effective patient categorization and treatment outcomes under constraints.