Reconstruction d’images en tomographie photoacoustique avec régularisation combinée variation totale - Cauchy
Date:
I presented a poster on our recent work at the joint workshop on proxy models and co-design for innovative systems. This event brings together researchers from the GDR Ondes, RT MAIAGES, and GDR IASIS communities to discuss advances in computational imaging.
The poster summarizes the key findings from our GRETSI 2025 paper, focusing on our combined Total Variation and Cauchy regularization approach for photoacoustic tomography. I particularly highlight how this method fits into a co-design framework by using efficient, on-the-fly matrix-vector products. This allows our advanced reconstruction algorithm to handle the large-scale data from the PAT scanner without requiring prohibitive amounts of memory, making it practical for real-world applications.