Starting in 2015, the Medical Imaging Multimodality Breast Cancer Diagnosis User Interface (MIMBCD-UI) initiative involves the collaborative effort of three Portuguese research institutions: ISR-Lisboa/LARSyS, ITI/LARSyS, and INESC-ID. These laboratories are Associate Laboratories of Instituto Superior Técnico (IST), part of the University of Lisbon (ULisboa), Portugal (EU). ISR-Lisboa/LARSyS and ITI/LARSyS are both part of the LARSyS research network, with ISR-Lisboa/LARSyS focusing on systems and robotics, while ITI/LARSyS is focusing on Human–Computer Interaction (HCI). Together, they bring complementary expertise to the MIMBCD-UI initiative, combining technical innovation with user-centered design.
The MIMBCD-UI initiative is the precursor of both MIDA and BreastScreening initiatives. In 2022, the team was funded by a national FCT project, namely Multiple Instance Attention Learning for Multimodal Breast Cancer (MIA-BREAST) with the FCT reference 2022.04485.PTDC (DOI: 10.54499/2022.04485.PTDC), receiving contributions from the MIMBCD-UI initiative.
In 2025, the team advanced this research through a new national FCT project entitled Integration of an Artificial Intelligence Agent in Radiology to Assist in Breast Cancer Diagnosis with the reference 2024.07344.IACDC (DOI: 10.54499/2024.07344.IACDC). This project, developed in collaboration with the Centro Hospitalar de Trás-os-Montes e Alto Douro (CHTMAD), focuses on implementing a pilot software solution to support breast cancer diagnosis by integrating MammoGraphy (MG), UltraSound (US), and Magnetic Resonance Imaging (MRI) modalities. The system leverages an AI agent to enhance radiologists' decision-making, reduce diagnostic errors, and improve workflow efficiency. The initiative includes explainability features, real-time diagnostic assistance, and training support for early-career radiologists, while also laying the foundation for a digital, data-centric hospital environment.
Additionally, this initiative was an important part of the BreastScreening-AI endeavor developments, a spinoff from IST. The knowledge, prototypes, and user interface designs developed under MIMBCD-UI directly informed the clinical and technical foundations of the BreastScreening-AI project (SensiPerception, LDA). This continuity ensures that scientific rigor and human-centered design remain at the core of translational efforts aimed at deploying AI in real-world radiology workflows.
For further information, follow the public wiki of the MIMBCD-UI initiative. Moreover, we also provide a private wiki on the meta-private repository for team usage. Unfortunately, you need to be a member of our team to access the restricted information.
