ESTRO 2025 - Abstract Book

S3015

Physics - Image acquisition and processing

ESTRO 2025

Conclusion: Our optimized DTI protocol offers a clinically feasible approach for accurate axonal mapping in SRT planning. This sequence enables robust identification of organs at risk with a reduced imaging time.

Keywords: DTI, MRI, SRT

References: [1] PMID: 31326352. [2] PMID: 22014949. [3] PMID: 35249816. [4] PMID: 19301968.

2410

Digital Poster Diffusion and echo time differences can introduce discrepancies in observed Apparent Diffusion Coefficients between MRI-linacs and diagnostic scanners Kerstin Lagerstrand 1,2 , Emilia Palmer 3 , Minea Jokivouolle 4,5 , Faisal Mahmood 4,5 1 Department of Medical Radiation Sciences, Institution of Clinical Sciences, Gothenburg, Sweden. 2 Department of Medical Physics and Biomedical Engineering physics, Sahlgrenska University Hospital, Gothenburg, Sweden. 3 Department of Medical Radiation Physics and Nuclear Medicine, Institution of Clinical Sciences, Stockholm, Sweden. 4 Department of Clinical Research, University of Southern Denmark, Odense, Denmark. 5 Department of Oncology, Odense University Hospital, Odense, Denmark Purpose/Objective: Diffusion-weighted imaging (DWI) has shown promise as a valuable MRI-technique for tumor delineation in radiation therapy, demonstrating high contrast between tumors and surrounding tissues. DWI offers additional advantages since the Apparent Diffusion Coefficient (ADC) holds promise as a quantitative biomarker for treatment response assessment. DWI can be implemented in both diagnostic scanners (DSs) and MRI-linacs (MRLs). Typically, scanning is optimized to increase the image quality and minimize the acquisition time by e.g. reducing the echo time within the limits of the gradient performance and the time needed to sample the signal at a specified spatial resolution and sampling bandwidth. Material/Methods: Patients scheduled for pancreas (n=10) and abdominal (n=11) DWI in a DS and MRL were evaluated before radiation therapy and within one of the first three treatment fractions. One patient had two abdominal targets. Optimized DWI protocols were employed, utilizing the shortest available echo time to maximize signal-to-noise ratio and the highest achievable bmax to improve tumor-to-normal tissue contrast. The delineated gross tumor volume was extracted from a planning MR image (T2w for all patients, except T1w for one liver patient). Mean tumor ADCs and ADC DS -ADC MRL were compared between scanners using scatter and Bland Altman plots. Results: The final DWI sequences for the DS and MRL are presented in Fig. 1. As can be seen, both the diffusion and echo time were shorter in DS compared to MRL. This study aims to determine whether DSs and MRLs exhibit ADC differences of clinical importance for evaluation of pancreatic and other abdominal tumors.