Reference-Free Meta-Learning for Generalized Implicit Neural Representation in Efficient MRI Reconstruction

Implicit Neural Representation (INR) has emerged as a powerful paradigm for continuous MRI reconstruction. However, standard self-supervised INR requires time-consuming optimization from scratch for each scan, hindering clinical deployment. This work presents IPOD, a Reference-Free Meta-Learning framework designed to learn generalized parameter initializations for INR directly from undersampled data. Distinct from conventional meta-learning that relies on fully-sampled ground truth, IPOD operates in an inverse-problem-driven manner, leveraging diverse reconstruction tasks with varying sampling patterns to capture a robust prior. Furthermore, we introduce an adaptive meta-update strategy modulated by task-specific performance to ensure optimal parameter distribution for diverse anatomical structures. Extensive experiments demonstrate that IPOD provides a superior initialization that enables rapid adaptation and achieves high-fidelity reconstruction across various imaging protocols, significantly outperforming existing INR baselines. By eliminating the dependence on reference images, IPOD offers a scalable and efficient solution for a wide range of imaging inverse problems. Code and data available at: https://github.com/zhn00310/RFML4MRI

(1) Problem: Magnetic Resonance Imaging (MRI) is a vital tool for medical diagnosis, but taking a full scan is often painfully slow for patients. To speed up the process, hospitals collect only a fraction of the raw data, but using standard computer methods to fill in the missing pieces requires starting a time-consuming calculation from scratch for every single patient. (2) Solution: To solve this, we developed a smart AI program named IPOD that builds an "imaging intuition" before meeting new patients. By practicing on a wide variety of quick, incomplete scans from past cases, the program learns how human anatomy generally looks under different scanning conditions. When a new patient arrives, IPOD uses this pre-learned intuition to rapidly piece together a crystal-clear image. (3) Impact: Crucially, unlike traditional methods, IPOD trains itself directly on the fast, incomplete scans that hospitals already collect, rather than requiring rare and flawless full reference images. This eliminates the long computing wait in clinics, making fast, high-quality MRI scans much more practical and accessible for everyday hospital environments.