Leidi, M.; Jia, Y.; Helbing, D.; Barranco, J.; Acikgoz, B. C.; Peper, E.; Ledoux, J.-B.; Bastiaansen, J. A. M.; Franceschiello, B.; Milani, B.

Magnetic Resonance Imaging (MRI) is a pivotal tool in modern diagnostics and research, yet image reconstruction is often used by clinicians and researchers as a black box. In this work, we introduce Monalisa, an open-source, user-friendly MATLAB framework designed to simplify the MRI reconstruction process. Monalisa decomposes the reconstruction pipeline into clear modular steps, including raw data reading with flexible file-type abstraction, trajectory computation, density compensation, advanced coil sensitivity mapping, and tailored binning strategies through its mitosius preprocessing stage. The framework supports a suite of reconstruction methods, including iterative-SENSE (also named CG-SENSE), GeneRalized Autocalibrating Partial Parallel Acquisition (GRAPPA) reconstructions, and Compressed Sensing (CS) supporting both spatial and temporal regularization using l1 and l2 techniques, accommodating both Cartesian and non-Cartesian acquisitions. Benchmark experiments comparing Monalisa with established frameworks such as Berkeley Advanced Reconstruction Toolbox (BART) on simulated 2D radial acquisitions demonstrate competitive performance, with trade-offs observed between Structural Similarity Index (SSIM) and l2 error metrics. In some benchmark reconstructions, Monalisa produced images with fewer visible artifacts than BART, although these differences were not always reflected in SSIM or l2 distance metrics. By providing comprehensive documentation and modular design, Monalisa serves not only as a powerful tool for research and clinical imaging but also as an educational platform to facilitate innovation in MRI reconstruction.