Needle localization in MR-guided biopsy and aspiration: effects of field strength, sequence design, and magnetic...

The purpose of this investigation was to evaluate the accuracy of MR Imaging for needle depiction at 0.2 and 1.5 T with multiple pulse sequences and needle orientations. The goal was to provide a framework for biopsy approach and imaging technique parameter selection that will ensure the safety and accuracy of MR-guided procedures.

Eight titanium and stainless steel alloy MR-compatible biopsy devices were immersed in fluid phantoms and placed into 1.5- and 0.2-T MR systems used for clinical imaging. Spin-echo, turbo spin-echo, and gradient-echo images were obtained with the needle shafts of the biopsy devices placed parallel to, perpendicular to, and at angles of 30 degrees and 60 degrees relative to the static magnetic field of the scanner. All images were obtained with the frequency-encoding direction parallel to and perpendicular to the needle shaft. Needle width and tip position were measured from images on a freestanding workstation, and the apparent tip position was compared with that obtained by direct measurement. The difference between these values was calculated for each needle type, imaging sequence, frequency-encoding direction, and needle orientation.

Artifactual widening was much more apparent at 1.5 T than at 0.2 T, as was error in determining needle tip position. Artifacts at both field strengths were most pronounced with gradient-echo sequences, less so with turbo spin-echo sequences, and least of all with spin-echo sequences. For spin-echo and turbo spin-echo sequences, when the frequency-encoding axis was perpendicular to the needle shaft, the apparent width of the needle was larger, but error in needle tip position was smaller. Artifacts were much less apparent, but error in tip position increased, as the orientation of the needle shaft became more parallel to the direction of the magnetic field.

Specific measurements differed with field strength, but needle tip localization within 1 mm was obtained at both 0.2 and 1.5 T with the appropriate frequency-encoding direction, pulse sequence, and imaging parameters. Orientation of the needle parallel to the magnetic field significantly reduced the apparent width of the needle at both field strengths but also decreased the accuracy of needle tip position localization.