Identifying treatment response in high-risk Neuroblastoma

Aim: To develop a classifier that can aid in the identification of patients that respond to current therapy in high-risk Neuroblastoma. Methods: Simulated Annealing (SA) in combination with the Support Vector Machine classification algorithm was applied to microarray data to identify a subset of genes whose expression levels allow the prediction of outcome upon megatherapy in high-risk patients. The training set and test set consisted of 40 patients each. In the training set, favourable and unfavourable outcome was defined as event-free survival of at least 3 years or event within 3 years after diagnosis, respec…