ESTRO 2023 - Abstract Book

S1850

Digital Posters

ESTRO 2023

20 patients were included in the study, 5 patients (25%) showed pCR. Bland-Altman analysis indicated close agreements of ADC obtained between both delineations with a non-significant bias. The delineations of the manual and semi-automatic method are displayed in Figure 1. The weekly ADCmean increase for the patients with a pCR was 22.5% for the semi- automatic segmentation and 16.0% for the manual segmentation respectively, whereas for the non-pCRs this was 8.4% and 8.8% (Figure 2). This increased distinctiveness for semi-automated method was also reflected in an increase of the AUC (0.84 vs 0.80 for the semiautomatic and manual workflow respectively).

Conclusion We have developed a semi-automated workflow that accurately and robustly segments the tumor volumes on multiple DW- MRI scans based on a single seed placement on the baseline scan. Analysis revealed similar or possibly even improved predictive performance compared to the manual method allowing this method for widespread application in larger patient cohorts and/or in centers with limited experience with DWI.

PO-2072 Simultaneous PET and DW-MR for Outcome Prediction of Neoadjuvant Chemoradiotherapy for Rectal Cancer

J. Wyatt 1 , G. Petrides 2 , R. Pearson 3 , A. Hashmi 3 , T. Simmons 4 , R. Maxwell 5 , H. McCallum 5

1 Newcastle University, Institute of Translational and Clinical Research, Newcastle upon Tyne, United Kingdom; 2 Newcastle upon Tyne Hospitals NHS Foundation Trust, Nuclear Medicine, Newcastle upon Tyne, United Kingdom; 3 Newcastle upon Tyne Hospitals NHS Foundation Trust, Northern Centre for Cancer Care, Newcastle upon Tyne, United Kingdom; 4 Newcastle upon Tyne Hospitals NHS Foundation Trust, Northern Centre for Cancer Care, Newcastle upon Tyne Hospitals NHS Foundation Trust, United Kingdom; 5 Newcastle University, Institute for Translational and Clinical Research, Newcastle upon Tyne, United Kingdom

Purpose or Objective