Pediatric Abdominal Scan at 3 T: SAR-Limited T2 Strategy

SAR in FSE scales approximately with the square of the refocusing flip angle and with the number of refocusing pulses per unit time (i.e., inversely with echo spacing and directly with ETL duty cycle). Reducing averages from 4 to 1 cuts scan time and total energy deposited over the exam, but the instantaneous SAR that triggered the abort is largely unchanged because each TR interval still contains the same number of refocusing pulses at the same flip angle. Worse, dropping to 1 average at 3 T in a 28 kg patient will leave SNR marginal for a small lesion.

High-yield levers, in order of efficiency:

1. Lower refocusing flip angle (e.g., 120 deg to 90-100 deg, or use hyperecho/variable flip trains). SAR falls roughly with flip-angle squared, so 120 to 90 deg is a ~44% reduction in RF power per pulse.
2. Increase echo spacing (fewer refocusing pulses per TR), accepting more T2 decay across the train and slightly more blurring.
3. Lengthen TR / reduce duty cycle if trigger timing allows.
4. Reduce the number of slices or use a 2D instead of 3D acquisition where coverage permits, since 3D FSE deposits RF across a larger volume per TR.

A sensible plan: keep at least 2-3 averages, drop the refocusing flip to ~90-100 deg with a variable-flip scheme, increase echo spacing modestly, and retain the single-shot T2 (low SAR, fast) as the fluid-contrast workhorse. Verify with the scanner's SAR prediction before scanning and monitor.

Respiratory triggering ties TR to the breathing cycle (typically 3-5 s in a child), which actually lowers average SAR because the RF duty cycle per unit time falls. The catch is that the scanner's predicted SAR is usually computed at the nominal TR; if triggering stretches TR, real-time SAR is lower than predicted, so the abort may be conservative. Conversely, if the sequence is run with a short fixed TR to finish faster, duty cycle rises and SAR climbs.

Given sedation and no breath-hold:

• Use respiratory triggering or navigator gating with a generous acceptance window rather than breath-hold, accepting a longer scan.
• Keep the single-shot T2 FSE as the primary fluid-sensitive sequence: one shot per trigger, very low SAR, motion-robust.
• For the 3D FSE, reduce flip angle and averages as above and let the trigger define TR; this both controls SAR and suppresses respiratory ghosting.
• If time is critical, consider a free-breathing radial or stack-of-stars T2 acquisition, which tolerates motion and typically uses lower flip angles.

Recheck predicted SAR against the 4 W/kg first-level limit and the local institutional policy for pediatric patients (often stricter).

SNR in FSE falls with lower refocusing flip angle because less transverse magnetization is refocused each echo. Recovery options that do not primarily add RF power:

• Increase voxel size modestly (e.g., 1.5 to 2 mm in-plane) - SNR scales with voxel volume; often acceptable for lesion detection.
• Use a body or torso phased-array coil with the smallest feasible geometry and combine channels optimally; surface-coil proximity is the cheapest SNR gain.
• Signal averaging adds SNR as sqrt(N) but also adds RF energy per unit time only if TR is fixed; with triggered TR, adding averages lengthens the exam rather than raising instantaneous SAR - acceptable if time allows.
• Reduce bandwidth (narrower receive bandwidth) to cut noise, at the cost of more chemical-shift and susceptibility artifact.
• Shorten the echo train so fewer low-flip-angle refocusing pulses degrade the signal, and use partial Fourier in the phase direction to shorten the train.

Avoid simply raising the flip angle back up; that re-creates the SAR problem. Document the trade-offs and confirm final image quality before the patient leaves the scanner.