Scalable Bayesian uncertainty quantification with learned convex regularisers for radio interferometric imaging
Tobias Liaudat, Irfu/Dedip
Vendredi 31 mai à 11h30, CEA-Saclay, Orme-des merisiers, Bat 709, salle Cassini
The last decade brought us substantial progress in computational imaging techniques for current and next-generation interferometric telescopes, such as the SKA. Imaging methods have exploited sparsity and more recent deep learning architectures with remarkable results. Despite good reconstruction quality, obtaining reliable uncertainty quantification (UQ) remains a common pitfall of most imaging methods. The UQ problem can be addressed by reformulating the inverse problem in the Bayesian framework. The posterior probability density function provides a comprehensive understanding of the uncertainties. However, computing the posterior in high-dimensional settings is an extremely challenging task. Posterior probabilities are often computed with sampling techniques, but these cannot yet cope with the high-dimensional settings from radio imaging.
This work proposes a method to address uncertainty quantification in radio-interferometric imaging with data-driven (learned) priors for very high-dimensional settings. Our model uses an analytic physically motivated model for the likelihood and exploits a data-driven prior learned from data. The proposed prior can encode complex information learned implicitly from training data and improves results from handcrafted priors (e.g., wavelet-based sparsity-promoting priors). We exploit recent advances in neural-network-based convex regularisers for the prior that allow us to ensure the log-concavity of the posterior while still being expressive. We leverage probability concentration phenomena of log-concave posterior functions that let us obtain information about the posterior avoiding the use of sampling techniques. Our method only requires the maximum-a-posteriori (MAP) estimation and evaluations of the likelihood and prior potentials. We rely on convex optimisation methods to compute the MAP estimation, which are known to be much faster and better scale with dimension than sampling strategies. The proposed method allows us to compute local credible intervals, i.e., Bayesian error bars, and perform hypothesis testing of structure on the reconstructed image. We demonstrate our method by reconstructing simulated radio-interferometric images and carrying out fast and scalable uncertainty quantification.
Rôles de l'IA dans le cancer du sein pour le dépistage à partir de mammographies et pour la prédiction de la réponse à la chimiothérapie néoadjuvante à partir d’IRM
Frédérique Frouin, institut curie
Mardi 30 avril à 11h30, CEA-Saclay, Orme-des merisiers, Bat 709, salle Cassini
L’exposé portera dans un premier temps sur l’utilisation possible des méthodes d’intelligence artificielle pour le dépistage du cancer du sein à partir de mammographies, domaine pour lequel les données d’apprentissage sont nombreuses. Dans un second temps, à partir de notre expérience de recherche, nous montrerons comment dans des domaines beaucoup plus restreints, nous avons pu utiliser des outils d’IA pour tenter de mieux prédire la réponse à la chimiothérapie néoadjuvante et nous identifierons un certain nombre de difficultés qui limitent l’impact de ces outils à l’heure actuelle.
