Contributions
Abstract: P1046
Type: Poster
Abstract Category: Pathology and pathogenesis of MS - Imaging
Background: White matter lesion (WML) central veins seen on MRI may be a useful imaging biomarker for the diagnosis of multiple sclerosis (MS). Various T2*-weighted sequences, including susceptibility weighted imaging (SWI) have been proposed to best delineate WMLs and central veins, but quantitative comparisons are lacking. Identifying a subset of WMLs with central veins will also ease clinical translation.
Objectives: We assessed the accuracy of diagnosis of MS and ischaemic small vessel disease (SVD) when using T2* and a fusion of fluid attenuated inversion recovery (FLAIR) and SWI (FLAIR-SWI) at 3T.
Method: 23 patients with RRMS and 16 with SVD were scanned on a 3T Philips Achieva MR scanner. T2*-weighted imaging with a high EPI factor and FLAIR imaging were acquired for each patient. SWI was produced by combination of the phase mask and magnitude images. FLAIR-SWI was produced by a fusion of FLAIR and SWI. Two blinded raters made a diagnosis of MS or SVD based only on counting a subset of WMLs with central veins using T2* and then FLAIR-SWI. Each rater repeated the process 1 week later. Only the presence of WML central veins was used in the radiological diagnosis of MS. Cohen´s kappa (ĸ) was used to determine a level of agreement of the radiological diagnosis based on the central vein sign and the actual clinical diagnosis. Sensitivities and specificities were calculated.
Results: Both raters, without any access to clinical details had good agreements with the clinical diagnosis irrespective of the vein imaging sequence. Rater 1 at two separate time points; (T2* ĸ=0.79 and 0.84) and (FLAIR-SWI ĸ=0.79 and 0.73), sensitivity = 91.3% and specificity ranging from 87.5-93.8% with T2* and 81.3 - 87.5% with FLAIR-SWI. Rater 2 at two separate time points; (T2* ĸ=0.65 and 0.73) and (FLAIR-SWI ĸ=0.61 and 0.84), sensitivity = up to 91.3% and specificity ranging from 81.3-100% with T2* and 93.8-100% with FLAIR-SWI.
Conclusions: T2* is just as effective as FLAIR-SWI when making a diagnosis of MS using the central vein sign. Good-very good agreements with the actual diagnosis were achieved with both sequences. This may aid translation of the central vein sign, especially for smaller hospitals that may not have the expertise to produce FLAIR-SWI.
Disclosure: Margareta Clarke: nothing to disclose; Amal PR Samaraweera: nothing to disclose; Olivier Mougin: nothing to disclose; Matthew South: nothing to disclose; Rob A Dineen: nothing to disclose; Paul S Morgan: nothing to disclose; Nikos Evangelou: nothing to disclose
Abstract: P1046
Type: Poster
Abstract Category: Pathology and pathogenesis of MS - Imaging
Background: White matter lesion (WML) central veins seen on MRI may be a useful imaging biomarker for the diagnosis of multiple sclerosis (MS). Various T2*-weighted sequences, including susceptibility weighted imaging (SWI) have been proposed to best delineate WMLs and central veins, but quantitative comparisons are lacking. Identifying a subset of WMLs with central veins will also ease clinical translation.
Objectives: We assessed the accuracy of diagnosis of MS and ischaemic small vessel disease (SVD) when using T2* and a fusion of fluid attenuated inversion recovery (FLAIR) and SWI (FLAIR-SWI) at 3T.
Method: 23 patients with RRMS and 16 with SVD were scanned on a 3T Philips Achieva MR scanner. T2*-weighted imaging with a high EPI factor and FLAIR imaging were acquired for each patient. SWI was produced by combination of the phase mask and magnitude images. FLAIR-SWI was produced by a fusion of FLAIR and SWI. Two blinded raters made a diagnosis of MS or SVD based only on counting a subset of WMLs with central veins using T2* and then FLAIR-SWI. Each rater repeated the process 1 week later. Only the presence of WML central veins was used in the radiological diagnosis of MS. Cohen´s kappa (ĸ) was used to determine a level of agreement of the radiological diagnosis based on the central vein sign and the actual clinical diagnosis. Sensitivities and specificities were calculated.
Results: Both raters, without any access to clinical details had good agreements with the clinical diagnosis irrespective of the vein imaging sequence. Rater 1 at two separate time points; (T2* ĸ=0.79 and 0.84) and (FLAIR-SWI ĸ=0.79 and 0.73), sensitivity = 91.3% and specificity ranging from 87.5-93.8% with T2* and 81.3 - 87.5% with FLAIR-SWI. Rater 2 at two separate time points; (T2* ĸ=0.65 and 0.73) and (FLAIR-SWI ĸ=0.61 and 0.84), sensitivity = up to 91.3% and specificity ranging from 81.3-100% with T2* and 93.8-100% with FLAIR-SWI.
Conclusions: T2* is just as effective as FLAIR-SWI when making a diagnosis of MS using the central vein sign. Good-very good agreements with the actual diagnosis were achieved with both sequences. This may aid translation of the central vein sign, especially for smaller hospitals that may not have the expertise to produce FLAIR-SWI.
Disclosure: Margareta Clarke: nothing to disclose; Amal PR Samaraweera: nothing to disclose; Olivier Mougin: nothing to disclose; Matthew South: nothing to disclose; Rob A Dineen: nothing to disclose; Paul S Morgan: nothing to disclose; Nikos Evangelou: nothing to disclose