Abstract #0727

Radiomic features on T2w MRI to predict tumor invasiveness for pre-operative planning in colorectal cancer: preliminary results

Jacob Antunes¹, Scott Steele², Conor Delaney², Joseph Willis³, Justin Brady⁴, Rajmohan Paspulati⁵, Anant Madabhushi¹, and Satish Viswanath¹

¹Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States, ²Department of Colon and Rectal Surgery, University Hospitals Case Medical Center, Cleveland, OH, United States, ³Department of Anatomic Pathology, University Hospitals Case Medical Center, Cleveland, OH, United States, ⁴Department of General Surgery, University Hospitals Case Medical Center, Cleveland, OH, United States, ⁵Department of Radiology, University Hospitals Case Medical Center, Cleveland, OH, United States

Pre-operative planning in colorectal cancer is highly dependent on extent of tumor into the mesorectum, but tumor margin is currently only assessed on excised pathology. Radiomic features may capture subtle microarchitectural changes on a restaging MRI, enabling characterization of tumor extent prior to surgery, even when residual disease may not be visually discernible. We present preliminary results for identifying radiomic features which discriminate invasive from noninvasive tumor on a 3 Tesla restaging T2w MRI in colorectal cancer. In a cohort of 24 patients, multi-scale gradient (Gabor) radiomic features demonstrated high accuracy in segregating patients with invasive colorectal cancer.

This abstract and the presentation materials are available to members only; a login is required.

Join Here