High-quality MRI reconstruction is essential for accurate clinical diagnosis. However, clinicians often face challenges with phase singularities, which create "poles" or artifacts in reconstructed images. A recent study introduced a breakthrough method for MRI phase pole correction using regularized nonlinear inversion (NLINV). This technique effectively removes these errors, ensuring smoother coil sensitivity maps and clearer imaging outcomes.

Advancing MRI Phase Pole Correction with NLINV

The researchers integrated detection and correction algorithms directly into the NLINV framework. Specifically, they utilized the Gauss-Newton method with iterative regularization. The algorithm detects phase poles in individual coil sensitivity maps by computing the curl in each pixel. Furthermore, it calculates a weighted average of the curl across all coils to ensure precision. Consequently, this process generates images free from the distracting singularities that often plague auto-calibrated sensitivities. Moreover, the team demonstrated that the algorithm also successfully removes phase poles in ESPIRiT coil sensitivity maps.

Notably, the algorithm performs exceptionally well in demanding clinical scenarios. The study evaluated its efficacy using accelerated brain scans and real-time cardiac MRI. As a result, the system maintained high image integrity even with very small auto-calibration regions. This efficiency is particularly beneficial for interactive radial MRI of the human heart, where speed and precision are vital. Consequently, radiologists and cardiologists can expect more reliable estimations of coil sensitivity profiles, leading to fewer misinterpretations during diagnosis.

Frequently Asked Questions

What are phase poles in MRI?

Phase poles are singularities that occur during image reconstruction. They often stem from ambiguities in the estimation of coil sensitivities, resulting in localized image artifacts that can obscure clinical details.

How does the new NLINV method improve imaging?

The method identifies these poles by calculating the magnetic curl within sensitivity maps. It then applies a corrective global optimization step within the reconstruction algorithm to eliminate them, producing smoother and more accurate images even from minimal data.

Can this method be used for different organs?

Yes, the researchers successfully evaluated this method for both accelerated brain imaging (MPRAGE) and real-time interactive imaging of the human heart, suggesting broad applicability across various MRI protocols.

Disclaimer: This content is for informational and educational purposes only. It does not constitute medical advice or a professional recommendation. Always consult with a qualified healthcare provider for diagnosis and treatment. Refer to the latest local and national guidelines for clinical practice.

References

Blumenthal M et al. Phase-Pole-Free Images and Smooth Coil Sensitivity Maps by Regularized Nonlinear Inversion. Magn Reson Med. 2026 Mar 12. doi: 10.1002/mrm.70333. PMID: 41820227.

Uecker M et al. Image reconstruction by regularized nonlinear inversion—Joint estimation of coil sensitivities and image content. Magn Reson Med. 2008;60(3):674-682. doi: 10.1002/mrm.21691.

Blumenthal M, Uecker M. Phase-Pole-Free Images and Smooth Coil Sensitivity Maps by Regularized Nonlinear Inversion. arXiv:2508.04685 [physics.med-ph]. 2025.