ESTRO 2024 - Abstract Book

S3803

Physics - Image acquisition and processing

ESTRO 2024

primarily pertaining to address metal artifacts and to ensure anatomical correspondence (between MRI and paired CT). The parameters considered as crucial for MR image quality were the magnetic field homogeneity and spatial integrity, while signal to noise ratio was considered less important. All developers performed multiple pre-processing steps, including spatial registration, table removal, resampling, intensity standardization, clip, and bias field correction. Concerning image quality and dose metrics, differences were observed according to the user’s and developer’s preferences. Common metrics for users included DVH, gamma index, Mean Error (ME), and Mean Absolute Error (MAE). In contrast, developers employed more complex metrics such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM). For dose metrics, the prevalent criteria for gamma analysis was a pass rate superior to 65% with a local 2%/2mm criterion with a threshold of 10%. However, a wide dispersion of thresholds was observed, ranging from 10% to 80%, for both PTV and OARs evaluation. (Figure)

Section 4 showed that half of the survey responders do not have a dedicated sCT QA program. Half of the responders also do not apply a sanity check before dose calculation. At least 60% apply a selection strategy before enrolling a patient in an MR-only workflow, with age (<18 years), weight (over- or under-weight) and presence of metal artifacts as main exclusion criteria. Finally, 95% of respondents stressed the need for vendor-neutral guidelines to standardize and homogenize the use of this technology in clinical practices.

Conclusion:

The results of this survey showed substantial variations in both the development, clinical use and assessment of sCT across European institutions. Both developers and clinical end-users confirmed our hypothesis that expert, vendor neutral, consensus guidelines are urgently needed and must focus on tools, metrics and clinical protocols for clinical sCT implementation.

Keywords: MRI-Only workflow, Survey Synthetic CT

774

Digital Poster