Multi-Task MPI real-collected FFL-based dataset

Deep learning, owing to its potent nonlinear fitting capacity, has been increasingly applied to magnetic particle imaging (MPI)-related tasks to enhance the system performance. Nevertheless, the efficacy of deep learning models is contingent upon the availability of extensive training data. Currently, the availability of real-collected datasets suitable for deep learning training and testing remains limited, thereby compromising the generalizability and practical efficacy of data-driven approaches in real-world deployment scenarios. To mitigate this limitation, the proposed MPI-image dataset supports multiple image-centric tasks including anisotropy analysis and denoising, thereby facilitating the development and validation of deep learning approaches under real-world conditions.