Background: Positron Emission Tomography ââ?¬â?? Computed Tomography (PET/CT) imaging is the basis for the\nevaluation of response-to-treatment of several oncological diseases. In practice, such evaluation is manually\nperformed by specialists, which is rather complex and time-consuming. Evaluation measures have been proposed,\nbut with questionable reliability. The usage of before and after-treatment image descriptors of the lesions for\ntreatment response evaluation is still a territory to be explored.\nMethods: In this project, Artificial Neural Network approaches were implemented to automatically assess treatment\nresponse of patients suffering from neuroendocrine tumors and Hodgkyn lymphoma, based on image features\nextracted from PET/CT.\nResults: The results show that the considered set of features allows for the achievement of very high classification\nperformances, especially when data is properly balanced.\nConclusions: After synthetic data generation and PCA-based dimensionality reduction to only two components,\nLVQNN assured classification accuracies of 100%, 100%, 96.3% and 100% regarding the 4 response-to-treatment\nclasses.
Loading....