Medical imaging plays a crucial role in the diagnosis and treatment of breast cancer, going beyond simple lesion visualization. New approaches studied at ESPCI Paris – PSL, in collaboration with medical teams, aim to extract finer information about tissue stiffness, organization, and vascularization. This data could help distinguish benign lesions from malignant tumors and guide medical interventions more precisely.
The MUSCAD project, led by Béatrice Berthon at the Physics for Medicine Paris Institute, aims to combine multiple pieces of information about a tumor – mechanical properties, vascularization, tissue organization – within a single three-dimensional ultrasound exploration. Machine learning methods will be used to analyze this data for better tumor characterization. This approach builds on previous research in shear wave elastography and aims to reduce unnecessary biopsies through clinical studies with the Institut Curie.
Meanwhile, at the Institut Langevin, Alexandre Aubry is exploring the speed of sound propagation in breast tissue with the DIONISOS project. The goal is to determine if variations in sound speed can provide additional information about lesion malignancy. This work is part of a broader effort to identify measurable physical quantities for a better understanding of the observed tissue in imaging.
Finally, a preclinical study conducted by the Institut Langevin, Gustave Roussy, and Aquyre Biosciences (in 2023) explored a rapid analysis method for breast surgical samples. Combining label-free optical microscopy and machine learning, this approach, to which Olivier Thouvenin contributed, showed potential for better identifying tumor margins during surgery and reducing re-interventions. Validation on a larger and more diverse set of samples is needed.
