Basic Y-90 Image-Based Dosimetry Workshop: Past and Present
Description
Interactive, educational, hands-on session (60 min total). The first part focuses on participant interaction with the web demo; the second part provides an expert synthesis of results. Emphasis is on building intuition rather than promoting a single workflow. Timing is flexible and may be adjusted on site to match engagement and room logistics. Participants should bring a smartphone, tablet, or laptop. Attendees declare their clinical practice and tools, for example Tc99m MAA planar only, planar plus SPECT/CT, no post-treatment imaging, or post-treatment Y90 Bremsstrahlung SPECT/CT or PET/CT(MR). Using a lightweight web demo, they run a guided mini-analysis on preloaded Y90 SIRT case. Participants first choose basic reconstruction parameters (iterations, filter, matrix) and see detection-focused image quality metrics update in real time (visual check, CNR, recovery coefficient). They then select a MIRD-based analysis path, from single or multicompartment to voxel dosimetry, which enables calculation of dose metrics such as D90, Dmean, and DVH coverage. With the model selected, the platform displays deviation from ground truth to show how choices across acquisition, reconstruction, quantitation, and dosimetry propagate error. Live polling and brief volunteer readouts surface the largest uncertainty drivers under different resource scenarios. An expert-guided summary consolidates the crowd results into a ranked error budget, identifies parameters that most influence dose accuracy for each scenario, and provides a minimal viable workflow for available clinical imaging tools. Attendees receive a simple decision tree for selecting dosimetry model complexity, suggested tolerance bands for plan-versus-verify checks, and a short QA checklist.