A prepolarized MRI (PMRI) scanner was used to image near metal implants in agar gel phantoms and in in vivo human wrists. Comparison images were made on 1.5-and 0.5-T conventional whole-body systems. The PMRI experiments were performed in a smaller bore system tailored to extremity imaging with a prepolarization magnetic field of 0.4 T and a readout magnetic field of 27-54 mT (1.1-2.2 MHz). Scan parameters were chosen with equal readout gradient strength over a given field of view and matrix size to allow unbiased evaluation of the benefits of lower readout frequency. Results exhibit substantial reduction in metal susceptibility artifacts under PMRI versus conventional scanners. A new artifact quantification technique is also presented, and phantom results confirm that susceptibility artifacts improve as expected with decreasing readout magnetic field using PMRI. This proof-of-concept study demonstrates that prepolarized techniques have the potential to provide diagnostic cross-sectional images for postoperative evaluation of patients with metal implants. Magn Reson Med 56:177-186, 2006.

Purpose: To examine peripheral nerve stimulation (PNS) thresholds for normal human subjects in magnetic resonance imaging (MRI) gradient coils, and determine if observed thresholds could be predicted based on gross physiologic measurements. Materials and Methods:PNS thresholds for 21 healthy normal subjects were measured using a whole-body gradient coil. Subjects were exposed to a trapezoidal echo-planar imaging (EPI) gradient waveform and the total change in gradient strength (⌬G) required to cause PNS as a function of the duration of the gradient switching time ( ) were measured. Correlation coefficients and corresponding P values were calculated for the PNS threshold measurements against simple physiologic measurements taken of the subjects, including weight, height, girth, and average body fat percentage, in order to determine if there were any easily observable dependencies. Results:No convincing correlations between threshold parameters and gross physiologic measurements were observed. Conclusion:These results suggest it is unlikely that a simple physiologic measurement of subject anatomy can be used to guide the operation of MRI scanners in a subjectspecific manner in order to increase gradient system performance while avoiding PNS. IN EXTREME MODES OF OPERATION, rapidly switched magnetic gradient coils used in magnetic resonance imaging (MRI) induce within-tissue electric fields of magnitudes and durations capable of exciting peripheral nerves. The occurrence of peripheral nerve stimulation (PNS) currently represents an upper limit to the operation of MRI gradient coils, and it is therefore important that all factors influencing PNS thresholds be investigated and understood.It has been demonstrated that there is large intersubject variability in stimulation thresholds for peripheral nerves, although individual thresholds are relatively consistent and are well described by simple twoparameter curves (1-4). The total change in gradient field (⌬G) required to cause stimulation increases linearly with the time, , over which the gradient is ramping (4):( 1 )The curve parameters SR min and ⌬G min describe limiting cases. ⌬G min is the limiting value of ⌬G when tends to zero; SR min is the minimum gradient slew rate required to induce nerve stimulation when tends to infinity. An echo-planar imaging (EPI) readout waveform will not result in stimulation unless the total change in gradient strength (⌬G) is greater than at least ⌬G min , regardless of the slew rate (SR) applied. Conversely, an EPI waveform will not stimulate unless the slew rate of the individual lobes is greater than at least SR min , regardless of what the lobe amplitudes are. Because inter-subject variations in PNS thresholds as expressed by Eq. [1] are large and consistent, there is both motivation and opportunity for the application of subject-specific operating limits in gradient coil systems.Our primary hypothesis was that one or more gross measurements of anatomic parameters (such as weight, height, or average body fat percentage) would cor...