CSSL-ISRVN: consistency self-supervised learning integrating ISTANet and sensitivity refinement-enhanced variational network for accelerated MRI reconstruction

Key Points

• To develop an advanced method for accelerating MRI reconstruction using limited data.
• Developed CSSL-ISRVN integrating sensitivity refinement and hybrid modeling.
• Implemented consistency self-supervision for enhanced reconstruction accuracy.
• Evaluated model performance on MRI datasets using simulation techniques.
• Achieved high-fidelity MRI images despite reduced sampling rates.
• Demonstrated robustness in reconstruction across various scenarios.