The recovery process for large burns can be complex and challenging due to several major issues like infections, fluid loss and electrolyte imbalance, long wound healing time, pain and its management, scarring and disfigurement. All these are leading to several long-term… Leggi tutto complications and the recovery process for large burns requires comprehensive and multidisciplinary care to address the various physical, emotional, and psychological challenges faced by patients. Medicated patches are used for long term treatment of large burns. Though this step is crittical in the healing progression, the evaluation of the effect ot the medical treatment is still done purely visually by the clinicians at the change of the patch every 12 or 24 hours. AIM aims to industrialize a smart patch enabling clinicians to objectively assess cellular-level healing progression induced by medical patches. It features a microstructured platform with an embedded optical window offering insights into healing tissue. Near Infrared reflection confocal microscopy captures images through the window. AI analysis segments tissue and cell components, providing clinicians with quantified therapy response. This approach enhances therapy optimization and personalized treatment for individual patient responses.

The population of black holes observed with gravitational-wave detectors is shaped by the astrophysical processes that drive their formation and evolution. As we accumulate observations, we become able to "reverse-engineer" the problem and better understand these processes. Scheduled for 2034,… Leggi tutto the space-based detector LISA will be sensitive to lower frequencies than current detectors, and is expected to detect a broad variety of systems, probing astrophysics in a wide range of scales. Unlike current detectors, there will constantly be gravitational-wave signals in the LISA datastream, which will overlap in time and frequency. Thus, standard population analysis methods will not apply. Glopop will propose a novel technique to perform population analysis that circumvents this issue, and fully take into account that the LISA datastream will be the gravitational-wave product of a single Universe. To do so, we will establish the link between different classes of LISA sources, designing an astrophysical model that outputs joined predictions, and use the most up-to-date deep learning methods to build a neural network and perform likelihood-free-inference. We will then demonstrate the capacity of our network on a mock LISA datastream generated with our astrophysical model, performing the first realistic population inference with LISA. This new approach will allow to fully exploit the scientific potential of LISA and our results will provide important knowledge for the preparation of the mission, increasing the fellow's visibility in the field. To achieve this ambitious goal, it is crucial that the fellow be placed in an institution that can supplement his experience on LISA data-analysis with knowledge on black hole astrophysics and machine-learning. UniMiB is therefore the ideal place, with recognised experts in both fields. This scientific program will be supplemented by training activities aiming to provide the fellow the skills to pursue a successful academic career.