Study of the spatiotemporal learning of dynamic positron emission tomography data for the improvement of diagnostic accuracy in breast cancer

This article is based on the talk delivered on 12 September 2022, in the Medical Physics section of the 108th National Congress of the Italian Physical Society (Milan). This work addresses the challenge of improving PET diagnostic accuracy through an alternative approach based on the analysis of time signal intensity patterns extracted from dynamic PET tissue time activity curves with a deep learning model. Our framework outperforms the discriminative potential of the classical SUV analysis, paving the way for more accurate dynamic PET-based lesion discrimination without additional acquisition time or invasive procedures. The full study has been published in IEEE Trans. Radiat. Plasma Med. Sci., 7 (2023) 630, with the title Spatiotemporal learning of dynamic positron emission tomography data improves diagnostic accuracy in breast cancer.