Project description
AI for radiology
Radiology relies on interpreting complex imaging alongside clinical information. AI could considerably advance the clinical workflow, but current AI tools perform only narrow tasks and cannot capture the full decision-making process. The ERC-funded SAGMA project aims to develop a general medical AI system that integrates multiple data types and supports real-world diagnostic reasoning. Researchers will first use large language models to extract structured information from radiology reports, enabling efficient training of specialised image analysis models. These models will then be coordinated by an AI agent that combines uncertainty estimates, clinical guidelines and laboratory data. Finally, the system will be tested in clinical workflows to assess its impact on diagnostic accuracy, efficiency and acceptance by radiologists.
Objective
Artificial intelligence (AI) holds immense potential for revolutionizing radiology by enhancing diagnostic accuracy, efficiency, and personalized patient care. However, current AI applications are limited to only specialized taskArtificial intelligence (AI) holds immense potential for revolutionizing radiology by enhancing diagnostic accuracy, efficiency, and personalized patient care. However, current AI applications are limited to only specialized tasks, thus failing to capture the complexity of radiological practice, which requires integrating multimodal data and nuanced decision-making. My vision with SAGMA is to develop an agent-based General Medical AI (GMAI) system that overcomes these limitations by combining specialized AI models with generalized reasoning capabilities. The project is structured around three central objectives:
1. Develop specialized image analysis AI models in a scalable manner by leveraging large language models (LLMs) to extract structured data from existing radiological reports. This will enable efficient training across various radiological tasks using unstructured clinical data. 2. Assemble a GMAI system using a LLM as an agent that coordinates the specialized AI models, incorporates uncertainty estimates, and integrates additional tools such as clinical guidelines and laboratory values. This system will utilize multimodal inputs to support comprehensive decision-making.
3.Validate the utility of the agent-based system in realistic clinical settings by assessing its impact on diagnostic accuracy, efficiency, and the overall radiological workflow, ensuring acceptance by clinical experts.
By achieving these objectives, SAGMA will bridge the gap between current AI capabilities and the multifaceted nature of radiological practice. The project will demonstrate how an agent-based GMAI system can augment and empower human expertise, potentially transforming radiology and paving the way for similar advancements in other medical specialties.
Fields of science (EuroSciVoc)
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques. See: The European Science Vocabulary.
CORDIS classifies projects with EuroSciVoc, a multilingual taxonomy of fields of science, through a semi-automatic process based on NLP techniques. See: The European Science Vocabulary.
You need to log in or register to use this function
Programme(s)
Multi-annual funding programmes that define the EU’s priorities for research and innovation.
Multi-annual funding programmes that define the EU’s priorities for research and innovation.
-
HORIZON.1.1 - European Research Council (ERC)
MAIN PROGRAMME
See all projects funded under this programme
Topic(s)
Calls for proposals are divided into topics. A topic defines a specific subject or area for which applicants can submit proposals. The description of a topic comprises its specific scope and the expected impact of the funded project.
Calls for proposals are divided into topics. A topic defines a specific subject or area for which applicants can submit proposals. The description of a topic comprises its specific scope and the expected impact of the funded project.
Funding Scheme
Funding scheme (or “Type of Action”) inside a programme with common features. It specifies: the scope of what is funded; the reimbursement rate; specific evaluation criteria to qualify for funding; and the use of simplified forms of costs like lump sums.
Funding scheme (or “Type of Action”) inside a programme with common features. It specifies: the scope of what is funded; the reimbursement rate; specific evaluation criteria to qualify for funding; and the use of simplified forms of costs like lump sums.
HORIZON-ERC - HORIZON ERC Grants
See all projects funded under this funding scheme
Call for proposal
Procedure for inviting applicants to submit project proposals, with the aim of receiving EU funding.
Procedure for inviting applicants to submit project proposals, with the aim of receiving EU funding.
(opens in new window) ERC-2025-STG
See all projects funded under this callHost institution
Net EU financial contribution. The sum of money that the participant receives, deducted by the EU contribution to its linked third party. It considers the distribution of the EU financial contribution between direct beneficiaries of the project and other types of participants, like third-party participants.
52074 Aachen
Germany
The total costs incurred by this organisation to participate in the project, including direct and indirect costs. This amount is a subset of the overall project budget.