TY - JOUR
T1 - Serial isotropic three-dimensional fast FLAIR imaging
T2 - Using image registration and subtraction to reveal active multiple sclerosis lesions
AU - Tan, I. Leng
AU - Van Schijndel, Ronald A.
AU - Pouwels, Petra J.W.
AU - Adèr, Herman J.
AU - Barkhof, Frederik
PY - 2002/1/1
Y1 - 2002/1/1
N2 - OBJECTIVE. Image registration and subtraction to detect the change of disease burden in multiple sclerosis on serial MR images should benefit from the use of high-resolution isotropic voxels. We compared 1.2-mm isotropic three-dimensional (3D) fast fluid-attenuated inversion recovery (FLAIR) images with standard 3-mm two-dimensional spin-echo images for the detection of new or enlarging lesions in longitudinal studies. SUBJECTS AND METHODS. Serial MR images were obtained at baseline, month 6 (n = 20), and month 7 (n = 16). For the half-yearly intervals, subtracted 3D FLAIR images and T2-weighted spin-echo images were compared. For the monthly intervals, subtracted 3D FLAIR images were compared with triple-dose contrast-enhanced T1-weighted spin-echo images. New, enlarging, and enhancing lesions were marked in consensus by two radiologists. RESULTS. At the half-yearly intervals, 3D FLAIR imaging detected more new or enlarging lesions than T2-weighted spin-echo imaging, both at the initial interpretation (80 vs 52; p < 0.001) and after a side-by-side comparison of the lesions (88 vs 65; p < 0.001). Post hoc analyses showed the largest benefit for new (rather than enlarging), for small, and for temporal lesions. At the monthly intervals, 32 enhancing lesions were detected on contrast-enhanced T1-weighted spin-echo images versus 20 new or enlarging lesions detected on 3D FLAIR images (p < 0.05). After a side-by-side comparison of the lesions, seven additional lesions were identified on 3D FLAIR images, making the difference with contrast-enhanced T1-weighted spin-echo images insignificant (27 vs 32; p > 0.05). CONCLUSION. Isotropic 3D FLAIR imaging holds great promise for the detection of new or enlarging lesions in multiple sclerosis using registration and subtraction techniques certainly at longer intervals.
AB - OBJECTIVE. Image registration and subtraction to detect the change of disease burden in multiple sclerosis on serial MR images should benefit from the use of high-resolution isotropic voxels. We compared 1.2-mm isotropic three-dimensional (3D) fast fluid-attenuated inversion recovery (FLAIR) images with standard 3-mm two-dimensional spin-echo images for the detection of new or enlarging lesions in longitudinal studies. SUBJECTS AND METHODS. Serial MR images were obtained at baseline, month 6 (n = 20), and month 7 (n = 16). For the half-yearly intervals, subtracted 3D FLAIR images and T2-weighted spin-echo images were compared. For the monthly intervals, subtracted 3D FLAIR images were compared with triple-dose contrast-enhanced T1-weighted spin-echo images. New, enlarging, and enhancing lesions were marked in consensus by two radiologists. RESULTS. At the half-yearly intervals, 3D FLAIR imaging detected more new or enlarging lesions than T2-weighted spin-echo imaging, both at the initial interpretation (80 vs 52; p < 0.001) and after a side-by-side comparison of the lesions (88 vs 65; p < 0.001). Post hoc analyses showed the largest benefit for new (rather than enlarging), for small, and for temporal lesions. At the monthly intervals, 32 enhancing lesions were detected on contrast-enhanced T1-weighted spin-echo images versus 20 new or enlarging lesions detected on 3D FLAIR images (p < 0.05). After a side-by-side comparison of the lesions, seven additional lesions were identified on 3D FLAIR images, making the difference with contrast-enhanced T1-weighted spin-echo images insignificant (27 vs 32; p > 0.05). CONCLUSION. Isotropic 3D FLAIR imaging holds great promise for the detection of new or enlarging lesions in multiple sclerosis using registration and subtraction techniques certainly at longer intervals.
UR - http://www.scopus.com/inward/record.url?scp=0036720996&partnerID=8YFLogxK
U2 - https://doi.org/10.2214/ajr.179.3.1790777
DO - https://doi.org/10.2214/ajr.179.3.1790777
M3 - Article
C2 - 12185063
SN - 0361-803X
VL - 179
SP - 777
EP - 782
JO - American Journal of Roentgenology
JF - American Journal of Roentgenology
IS - 3
ER -