Feature selection for outcome prediction in oesophageal cancer using genetic algorithm and random forest classifier

Outcome prediction is the foundation for tailoring and adapting a treatment planning in cancer therapy. Medical imaging plays a fundamental role in assessing the response to a treatment. In oncology, the Standard Uptake Value (SUV) of 18-FluoroDeoxyGlucose (FDG) measured by PET is widely used for diagnosing, staging and monitoring response to therapy Van De Wiele et al. (2013). Predictive (prediction of treatment response) and prognostic studies (prediction of the survival) using image features derived from 1st order statistics, such as Metabolic Tumour Volume (MTV) or Total Lesion Glycolysis (TLG: SUVmean × MTV), have been carried out.

Source: Computerized Medical Imaging and Graphics - December 27, 2016 Category: Radiology Authors: Desbordes Paul, Ruan Su, Modzelewski Romain, Vauclin S ébastien, Vera Pierre, Gardin Isabelle Source Type: research