PISCO Software v2.0

This software corresponds to a newer version of the original PISCO software, which is available at: http://mr.usc.edu/download/pisco/

Overview

This MATLAB software performs subspace-based estimation of sensitivity maps for multichannel MRI reconstruction. It follows a nullspace-based approach according to the theoretical framework and computational methods originally proposed in [1] and [2].

Key Improvements

The main differences with previous versions are:

1. Sketched SVD Method: A new computational method based on sketched SVD has been added. This method has been recently proposed in [3].

2. Expanded 2D/3D Data Support: The software now supports both 2D and 3D data processing.

3. Page-wise Matrix Operations: Functions based on page-wise matrix operations have been implemented to improve performance.

Contents

• Main Function: PISCO_sensitivity_map_estimation.m - Contains all the PISCO computational methods
• Example Files: example_2D.m and example_3D.m - Demonstration scripts for 2D and 3D data processing
• Documentation: PISCO_software_v2_0_documentation.pdf - Detailed documentation in PDF format

Getting Started

Please refer to the included PDF documentation and example files to begin using the software.

References

[1] R. A. Lobos, C.-C. Chan, J. P. Haldar. New Theory and Faster Computations for Subspace-Based Sensitivity Map Estimation in Multichannel MRI. IEEE Transactions on Medical Imaging 43:286-296, 2024.

[2] R. A. Lobos, C.-C. Chan, J. P. Haldar. Extended Version of "New Theory and Faster Computations for Subspace-Based Sensitivity Map Estimation in Multichannel MRI", 2023, arXiv:2302.13431. (https://arxiv.org/abs/2302.13431)

[3] R. A. Lobos, X. Wang, R. T. L. Fung, Y. He, D. Frey, D. Gupta, Z. Liu, J. A. Fessler, D. C. Noll. Spatiotemporal Maps for Dynamic MRI Reconstruction, 2025, arXiv:2507.14429. (https://arxiv.org/abs/2507.14429)