Rectal Cancer

Development of radiomic signatures in rectal cancer to predict CRT response.

Background:
Radiomics has been demonstrated to provide valuable information that are non-visible to the human eye on cancer patients. Tissue heterogeneity that is a hallmark feature of cancer has been shown to be assessed more accurately by means of various texture features not only on raw MR Images like T1 and T2, but also on parametric maps like ADC or Ktrans.

Objectives:
To develop and clinically apply models based on Machine Learning algorithms in patients with rectal cancer for prediction of treatment response before CRT.

Target Users:
Radiologists, Oncologists and Colorectal Surgeons.

Implications for Patient Care:
Unnecessary CRT can be avoided in patient groups that could not respond to such treatment, gaining valuable time for surgery, reducing the healthcare costs and improving quality of life.