PURPOSE: To describe the CT and MR characteristics of intraocular silicone oil (polydimethylsiloxane), which is used with increasing frequency to treat complicated retinal detachments. METHODS: CT was performed on a silicone oil/water phantom and on a patient with retinal detachments secondary to cytomegalovirus retinitis, treated by bilateral intraocular injections of silicone oil. CT appearance and CT number of silicone oil were evaluated. Proton MR spectroscopy was performed with a 200-MHz spectrometer on a sample of polydimethylsiloxane within a tube of deuterated water. MR imaging was performed on a silicone oil/water phantom and on two patients with retinal detachments treated with silicone oil injection. RESULTS: Silicone oil is relatively radiodense; its CT attenuation is approximately 130 HU. On spectroscopy, silicone oil gave a single peak at 0.33 ppm. Relative to water silicone oil was hyperintense on T1-weighted images and hypointense on spin-density and T2-weighted images. Estimated T1 and T2 were 716 msec and 68 msec, respectively. Chemical shift artifacts were seen on MR images and were exaggerated when a narrow sampling bandwidth was used. In clinical cases spectral saturation pulses normally used for lipid suppression could be adjusted to saturate only the silicone resonance; in this way, the chemical shift artifact was eliminated. CONCLUSION: Intraocular silicone oil has unique imaging characteristics with which radiologists must become familiar. These characteristics include high attenuation on CT and hyperintensity on T1-weighted MR, both of which may mimic hemorrhage. Elimination of the prominent chemical shift artifact on MR with selective saturation of the silicone resonance improves evaluation of the globe.
PURPOSE: To describe the CT and MR characteristics of intraocular silicone oil (polydimethylsiloxane), which is used with increasing frequency to treat complicated retinal detachments. METHODS: CT was performed on a silicone oil/water phantom and on a patient with retinal detachments secondary to cytomegalovirus retinitis, treated by bilateral intraocular injections of silicone oil. CT appearance and CT number of silicone oil were evaluated. Proton MR spectroscopy was performed with a 200-MHz spectrometer on a sample of polydimethylsiloxane within a tube of deuterated water. MR imaging was performed on a silicone oil/water phantom and on two patients with retinal detachments treated with silicone oil injection. RESULTS:Silicone oil is relatively radiodense; its CT attenuation is approximately 130 HU. On spectroscopy, silicone oil gave a single peak at 0.33 ppm. Relative to watersilicone oil was hyperintense on T1-weighted images and hypointense on spin-density and T2-weighted images. Estimated T1 and T2 were 716 msec and 68 msec, respectively. Chemical shift artifacts were seen on MR images and were exaggerated when a narrow sampling bandwidth was used. In clinical cases spectral saturation pulses normally used for lipid suppression could be adjusted to saturate only the silicone resonance; in this way, the chemical shift artifact was eliminated. CONCLUSION: Intraocular silicone oil has unique imaging characteristics with which radiologists must become familiar. These characteristics include high attenuation on CT and hyperintensity on T1-weighted MR, both of which may mimic hemorrhage. Elimination of the prominent chemical shift artifact on MR with selective saturation of the silicone resonance improves evaluation of the globe.