Prediction of the Response to Neoadjuvant Radiation Chemotherapy Through Texture Analysis Derived From Medical Imaging

Prediction of the Response to Neoadjuvant Radiation Chemotherapy Through Texture Analysis Derived From Magnetic Resonance Imaging, Contrast Injection Computed Tomography and Positron Emission Tomography at 18-FDG in Locally Advanced Rectal Cancer

In patient with complete pathological response, surgery could be omitted without compromising progression-free or overall survival. A powerful prediction model is needed to guide management of this population.

Radiomics features provide complementary information about tumor heterogenity.

The aim of the investigators is to develop a model combining clinical and radiomic criteria able to predict complete pathological response.

Inclusion Criteria:

• locally advanced rectal cancer
• patients who received concomitant radio-chemotherapy followed by surgery
• pre-therapeutic MRI and CT-scan dicom available

Exclusion Criteria:

• age under 18 yo
• lack of MRI or CT-scan