In the following paper, a method for dose reconstruction in a Boron Neutron Capture Therapy (BNCT) ideal facility is described. The BNCT combines a precise radiation beam with a targeted drug delivery, which makes the therapy particularly effective for unbeatable cancers like brain and melanoma tumours. A Single Photon Emission Computed Tomography set-up was developed to obtain the dose reconstruction. The measured data obtained are joined to ones obtained with the full simulated facility in Fluka. The dataset was used to train and validate a Deep Learning method for automatic dose reconstruction, obtaining an IoU = 91% and SSIM = 0.9 for the U-Net tight frame model.
BNCT-SPECT Deep Learning Reconstruction for an Online Dose Monitoring
Mazzucconi, Davide;Ferri, Tommaso;Grisoni, Luca;Ghisio, Filippo;Santini, Letizia;Borghi, Giacomo
2026-01-01
Abstract
In the following paper, a method for dose reconstruction in a Boron Neutron Capture Therapy (BNCT) ideal facility is described. The BNCT combines a precise radiation beam with a targeted drug delivery, which makes the therapy particularly effective for unbeatable cancers like brain and melanoma tumours. A Single Photon Emission Computed Tomography set-up was developed to obtain the dose reconstruction. The measured data obtained are joined to ones obtained with the full simulated facility in Fluka. The dataset was used to train and validate a Deep Learning method for automatic dose reconstruction, obtaining an IoU = 91% and SSIM = 0.9 for the U-Net tight frame model.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
