Recent advances in MRI gradient design and amplifiers have enabled clinical scanners to achieve very high slew rates, often several hundred T/m/s. However, high-resolution echo-planar imaging (EPI) protocols cannot routinely exploit full gradient performance because of limits imposed by peripheral nerve stimulation (PNS). The authors propose a pragmatic pulse-shape modification called Peripheral Nerve Stimulation Optimized Pulses for EPI (POPE) to characterize and mitigate those PNS constraints.
POPE is built on the premise that PNS risk is not uniform across an entire gradient waveform: brief epochs produce the highest predicted PNS spikes. Rather than reducing gradient performance globally, POPE selectively reduces the local slew rate at times when predicted PNS spikes would otherwise occur, while leaving the remainder of the waveform—and thus overall sequence performance—unchanged.
The POPE approach consists of identifying periods within the EPI gradient waveform that are predicted to generate high PNS responses and applying a targeted reduction in slew rate during those periods. The method therefore modifies only the pulse shape where necessary and preserves maximal gradient performance elsewhere in the sequence.
The authors describe POPE as a simple sequence modification that can be applied to EPI pulse waveforms. The abstract does not provide detailed algorithmic steps, parameter values for slew-rate reductions, or implementation code; those specifics were not reported in the provided source material. The central technical insight is selective, not wholesale, modulation of gradient slew to manage PNS exposure while maintaining high spatial resolution and fast acquisition.
Applying POPE to a range of EPI protocols produced reductions in acquisition time of approximately 7%–35% for resolutions spanning from 1.0 mm down to 0.3 mm. These improvements are reported as achievable without exceeding predicted PNS limits.
By avoiding a global slowdown of gradient waveforms, POPE realizes local improvements in effective imaging speed. The degree of improvement varies with the target voxel size: the abstract reports the 7%–35% range across the tested resolution set but does not break down gains by specific protocol details in the summary provided.
A key outcome highlighted by the authors is that POPE enables robust 0.3 mm isotropic fMRI protocols on clinical 7T MRI scanners. According to the abstract, such protocols would have exceeded safety limits for PNS without the POPE modification. By selectively shaping gradient pulses, POPE permits locally precise functional activation mapping at spatial resolutions that were previously limited by PNS constraints on clinical hardware.
The abstract frames this capability as facilitating high spatial specificity for fMRI activation mapping on standard clinical 7T systems, extending achievable resolutions for studies seeking fine-grain localization.
The motivation for POPE arises from the safety limits tied to peripheral nerve stimulation. The method explicitly aims to remain within predicted PNS thresholds while recovering otherwise unusable gradient performance. The abstract emphasizes that POPE achieves faster imaging while not exceeding predicted PNS, but it does not provide the numerical PNS thresholds, the prediction model used, or empirical PNS measurements in human subjects within the summary.
Readers should note that the abstract reports that POPE kept acquisitions within predicted PNS limits; details on how PNS prediction was validated, whether in vivo monitoring was performed, or exact safety margins are not reported in the provided text.
The work is authored by a multi-institutional team, primarily associated with the Martinos Center at MGH/Harvard Medical School, with collaborators from Siemens Healthineers, Maastricht University, Brain Innovation B.V., UC Berkeley, and the German Center for Neurodegenerative Diseases.
Funding support is declared from the National Institutes of Health (P41EB030006), and a data repository DOI is provided in the manuscript metadata. Competing interests disclosed in the abstract note that one author is an employee of Siemens Healthineers and another is an employee of Brain Innovation B.V.
The abstract does not include detailed methodological parameters, raw data, or implementation code; those materials may be available in the full manuscript or supplementary materials but were not detailed in the provided source excerpt.