PURPOSE: Interventional magnetic resonance (MR)-guided transcatheter embolization could potentially limit radiation exposure and improve visualization of target organs. The feasibility of monitoring injection and distribution of embolic agents was assessed in a dynamic flow model with real-time MR imaging. MATERIALS AND METHODS: MR-compatible flow models were constructed with use of clear plastic chambers containing 170-microm polyethylene tubular filters. Gadolinium (Gd)-impregnated polyvinyl alcohol (PVA) particles (355-500 and 500-710 microm in size) and Gd-impregnated microspheres (Embospheres, 300-500 and 500-700 microm in size) were injected into the flow circuit under real-time dynamic T1-weighted fast field echo guidance at four images per second. A dynamic steady-state free precession sequence at four images per second was used to monitor the injection of unmodified Embo-Gold 700-900- microm particles. High-resolution scans were obtained before and after each particle injection. RESULTS: MR signal enhancement on the dynamic T1-weighted fast field echo sequence was visible during the injection of Gd-impregnated microspheres. Gd-impregnated PVA particles were not detected by this sequence. After injection, microsphere and PVA localization to the filter chambers was confirmed by the high-resolution scans. On the high-resolution sequences, relative MR signal enhancement of the microspheres was higher than that of the PVA particles. The Embo-Gold particles were minimally detectable on the dynamic sequence and undetectable by the high-resolution scan. After particle injection, direct inspection of the filter chamber showed trapping of all particle types and sizes. CONCLUSION: Real-time MR tracking of Gd-impregnated embolic agents is possible in vitro.
PURPOSE: Interventional magnetic resonance (MR)-guided transcatheter embolization could potentially limit radiation exposure and improve visualization of target organs. The feasibility of monitoring injection and distribution of embolic agents was assessed in a dynamic flow model with real-time MR imaging. MATERIALS AND METHODS: MR-compatible flow models were constructed with use of clear plastic chambers containing 170-microm polyethylene tubular filters. Gadolinium (Gd)-impregnated polyvinyl alcohol (PVA) particles (355-500 and 500-710 microm in size) and Gd-impregnated microspheres (Embospheres, 300-500 and 500-700 microm in size) were injected into the flow circuit under real-time dynamic T1-weighted fast field echo guidance at four images per second. A dynamic steady-state free precession sequence at four images per second was used to monitor the injection of unmodified Embo-Gold 700-900- microm particles. High-resolution scans were obtained before and after each particle injection. RESULTS: MR signal enhancement on the dynamic T1-weighted fast field echo sequence was visible during the injection of Gd-impregnated microspheres. Gd-impregnated PVA particles were not detected by this sequence. After injection, microsphere and PVA localization to the filter chambers was confirmed by the high-resolution scans. On the high-resolution sequences, relative MR signal enhancement of the microspheres was higher than that of the PVA particles. The Embo-Gold particles were minimally detectable on the dynamic sequence and undetectable by the high-resolution scan. After particle injection, direct inspection of the filter chamber showed trapping of all particle types and sizes. CONCLUSION: Real-time MR tracking of Gd-impregnated embolic agents is possible in vitro.