The paper "ESPIRiT — An Eigenvalue Approach to Autocalibrating Parallel MRI: Where SENSE meets GRAPPA" by Martin Uecker et al. bridges the gap between two main approaches to parallel MRI: SENSE and GRAPPA. The authors analyze both methods as subspace methods, where SENSE uses precalculated sensitivity maps, while autocalibrating methods use calibrated kernels in k-space. They show that the dominant eigenvector of these k-space operators behaves like sensitivity maps and can be computed using an eigenvalue decomposition in image space. This leads to robust and high-quality sensitivity maps that can be estimated from autocalibration lines in k-space. The paper introduces a new method called ESPIRiT, which combines the benefits of both SENSE and GRAPPA. ESPIRiT uses multiple sets of sensitivity maps to enforce relaxed ("soft") sensitivity constraints in an extended SENSE-based reconstruction algorithm, offering robustness against certain types of errors similar to autocalibrated methods. The authors validate their approach through experimental examples, demonstrating its effectiveness in various scenarios, including reduced field-of-view (FOV) and noise levels. The results show that ESPIRiT can achieve better image quality than other methods, such as GRAPPA, and is more robust to errors and noise.The paper "ESPIRiT — An Eigenvalue Approach to Autocalibrating Parallel MRI: Where SENSE meets GRAPPA" by Martin Uecker et al. bridges the gap between two main approaches to parallel MRI: SENSE and GRAPPA. The authors analyze both methods as subspace methods, where SENSE uses precalculated sensitivity maps, while autocalibrating methods use calibrated kernels in k-space. They show that the dominant eigenvector of these k-space operators behaves like sensitivity maps and can be computed using an eigenvalue decomposition in image space. This leads to robust and high-quality sensitivity maps that can be estimated from autocalibration lines in k-space. The paper introduces a new method called ESPIRiT, which combines the benefits of both SENSE and GRAPPA. ESPIRiT uses multiple sets of sensitivity maps to enforce relaxed ("soft") sensitivity constraints in an extended SENSE-based reconstruction algorithm, offering robustness against certain types of errors similar to autocalibrated methods. The authors validate their approach through experimental examples, demonstrating its effectiveness in various scenarios, including reduced field-of-view (FOV) and noise levels. The results show that ESPIRiT can achieve better image quality than other methods, such as GRAPPA, and is more robust to errors and noise.