Cryoneedle Artifacts During MRI-Guided Cryoablation: Sources and Potential Mitigation Strategies

Purpose: Cryoneedle artifacts are frequently observed in MRI-guided cryoablations, and may obscure visualization of critical anatomy and compromise needle placement accuracy. This work experimentally investigated the contributing factors of these artifacts to identify effective mitigation strategies. Materials and Methods: Ex vivo porcine tissue with inserted cryoneedles was imaged on a 1.5-Tesla MRI. Fast spin echo (FSE) and spoiled gradient echo (GRE) sequences with echo times from 1.04 to 60 ms and specific absorption rates (SARs) from 0.01 to 2.1 W/kg were used. During MRI, cryoneedle temperatures were monitored using fiber-optic sensors. Configurations with one to three cryoneedles oriented at 0-degree or 45-degree angles to the patient table were investigated. The body coil was used for transmit/receive, both with and without an additional receive-only surface loop coil. Artifact width and intensity were measured for analysis. Results: Cryoneedle artifact widths were unrelated to echo time for both FSE (p = 0.6) and GRE (p = 0.3) and were smaller in GRE than in FSE images (p << 0.05). Artifact widths correlated with cryoneedle temperature elevations (r² = 0.969, p << 0.05) but were not correlated with SAR (GRE: p = 0.3; FSE: p = 0.5). The artifact intensity with the cryoneedle oriented at 0 degrees increased with a greater number of cryoneedles in the tissue (p = 0.006), and when the surface loop coil was used (p = 0.008). Conclusion: Clinically observed cryoneedle artifacts compromising treatment efficacy can be indicative of tissue radiofrequency heating risk, and effectively mitigated by either using GRE-based sequences or adjusting coil/cryoneedle configurations.