【摘要】 目的:通过对猪腰椎间关节软骨1.5T磁共振(magnetic resonance,MR)多序列成像与大体标本对照的实验性研究,评估MR成像(MR image,MRI)显示腰椎间关节软骨的能力与限度。方法:用Siemens1.5T超导磁共振对6条新鲜离体成年猪腰椎行椎间关节矢状面扫描。扫描序列包括:快速自旋回波T1加权序列(fast spin echo T1-weighted imaging,FSE T1WI)、快速自旋回波T2加权序列(fast spinecho T2-weighted imaging,FSE T2WI)、脂肪抑制质子加权序列(fat saturation proton density-weighted imaging,FS PDWI)、不同反转角的三维快速小角度激发成像序列(three-dimensional fast low angle shot imaging,3D-FLASH)以及不同反转角的水激发三维快速小角度激发成像序列(water excitation three-dimensional fast lowangle shot imaging,WE3D-FLASH)。扫描结束后将标本深低温冷冻,沿椎间关节矢状面正中层面切开,对照观察椎间关节矢状面正中层面大体标本与MRI图像。取6对L3/L4椎间关节矢状面正中层面各序列图像测量并计算软骨的信噪比(signal-noise ratio,SNR)以及软骨与骨皮质、软骨与骨髓、软骨与生理盐水的对比噪声比(contrast-noise ratio,CNR)。以游标卡尺测量6对L3/L4椎间关节软骨矢状面正中层面中心的厚度,与MRI所测量软骨厚度对照。结果:3D-FLASH(FA20°)及WE3D-FLASH(FA20°)序列显示腰椎间关节软骨及周围结构优于其他研究序列,相对FS PDWI序列能更真实地反映腰椎间关节软骨厚度。(1)4种不同反转角(FA10°,FA20°,FA30°及FA40°)3D-FLASH序列比较:3D-FLASH(FA20°)序列图像软骨SNR、软骨与骨皮质CNR最高,与其他3种3D-FLASH序列比较差异有统计学意义(P<0.05),软骨与骨髓CNR、软骨与生理盐水CNR亦较高。(2)4种不同反转角(FA10°,FA20°,FA30°及FA40°)WE3D-FLASH序列比较:WE3D-FLASH(FA20°)序列图像软骨SNR、软骨与骨皮质CNR、软骨与骨髓CNR最高,与其他3种WE3D-FLASH序列比较差异有统计学意义(P<0.05),软骨与生理盐水CNR亦较高。(3)3D-FLASH(FA20°),WE3D-FLASH(FA20°),FS PDWI,FSE T1WI及FEST2WI5种序列比较:3D-FLASH(FA20°)序列的软骨SNR、软骨与骨皮质CNR最高,与FS PDWI,FSET1WI及FSE T2WI序列比较差异有统计学意义(P<0.05),与WE3D-FLASH(FA20°)序列比较差异无统计学意义(P>0.05)。WE3D-FLASH(FA20°)序列软骨与骨髓CNR最高,与FS PDWI,FSE T1WI及FSE T2WI序列比较差异有统计学意义(P<0.05),与3D-FLASH(FA20°)序列比较差异无统计学意义(P>0.05)。FS PDWI和FSE T2WI序列软骨与生理盐水CNR高,两者间差异无统计学意义(P>0.05)。3D-FLASH(FA20°)和WE3D-FLASH(FA20°)序列软骨与生理盐水的CNR低,两者间差异无统计学意义(P>0.05);但两者分别与FS PDWI和FSE T2WI序列比较差异有统计学意义(P<0.05)。(4)3D-FLASH(FA20°)和WE3D-FLASH(FA20°)序列相对FS PDWI序列显示腰椎间关节软骨厚度更接近大体标本软骨测量厚度,前两者序列分别与后者比较差异均有统计学意义(P<0.05)。结论:3D-FLASH(FA20°)序列及其衍生的WE3D-FLASH(FA20°)序列能清晰地显示腰椎间关节软骨,且接近软骨的真实厚度,推荐作为腰椎间关节软骨检查的首选扫描序列。更多还原

【Abstract】 Objective To evaluate the capability and limitation of magnetic resonance image(MRI)for lumbar zygapophyseal joint cartilage through comparing pig lumbar zygapophyseal joint cartilage acquired from multiple MRI sequences of a 1.5 Tesla MR and gross specimens.Methods Six fresh lumbar spines from adult pigs were sagittaly scanned by Siemens 1.5 Tesla MR. The scan sequences included fast spin echo T1-weighted imaging (FSE T1WI),fast spin echo T2-weighted imaging (FSE T2 weighted T2WI),fat saturation proton density-weighted imaging (FS PDWI),3-dimensional fast low angle shot imaging (3D-FLASH),and water excitation 3-dimensional fast low angle shot imaging (WE 3D-FLASH). Each scan sequence acquired images from the same layer. The signal-noise ratio (SNR) for articular cartilage,contrast-noise ratio (CNR) for cartilage versus bone cortex,cartilage versus bone marrow,and cartilage versus saline were calculated.Right after the scanning,the lumbar spines were snap-frozen,incised sagittally along the midline lumbar zypapophyseal joints,and photographed to compare the gross specimens with corresponding MRIs. The thickness of sagittal midline center of 6 pairs of lumbar(L3/L4)zypapophyseal joint cartilage was measured by vernier caliper. The thickness of the back ventral articular cartilage was added and then compared with corresponding MR images.Results 3D-FLASH(FA 20°)and WE 3D-FLASH(FA 20°) sequences had significant advantages compared with other sequences in imaging lumbar zypapophyseal joint cartilage,and were mostly close to the real thickness.(1) Comparison of the 4 flip angle (FA10°,FA 20°,FA 30°,and FA40°) 3D-FLASH sequences:The highest cartilage SNR and best CNR of cartilage versus bone cortex were both found in the 3D-FLASH(FA 20°) sequence,which was significantly different from the other three 3D-FLASH sequences.The satisfacory CNR of cartilage versus bone marrow,cartilage versus saline were found more in the 3D-FLASH(FA 20°) sequence.(2) Comparison of the 4 flip angle(FA10°,FA 20°,FA 30°,and FA40°) WE 3D-FLASH sequences:The highest cartilage SNR,best CNR of cartilage versus bone cortex,and best CNR of cartilage versus bone marrow were found in the WE 3D-FLASH (FA 20°) sequence,which was significantly different from the other three 3D-FLASH sequences. The CNR of cartilage versus saline was found more satisfactory in the WE 3D-FLASH(FA 20°) sequence.(3) The highest cartilage SNR and best CNR of cartilage versus bone cortex were both found in the 3D-FLASH(FA 20°) sequence,which was significantly different from those in the PDWI,FSE T1WI,and FSE T2WI sequences (P<0.05),but with no significance (P>0.05) in the WE 3D-FLASH (FA 20°) sequence.The highest CNR of cartilage versus bone marrow was seen in WE 3D-FLASH(FA 20°) sequence. It was statistically significant compared with that in FS PDWI,FSE T1WI,and T2WI sequences respectively,but the difference was not significant compared with 3D-FLASH(FA 20°) sequence (P>0.05).Both the FS PDWI and T2WI sequences displayed ideal CNR of cartilage versus saline,with no significant difference (P>0.05). The lower SNR of cartilage versus saline was shown in 3D-FLASH(FA 20°) and WE 3D-FLASH(FA 20°) sequence,and the difference was not significant (P>0.05). However,they were significantly different compared with FS PDWI and T2WI sequences (P<0.05). (4) WE 3D-FLASH(FA20°) and 3D-FLASH(FA20°) sequences were relatively better than the FS PDWI when comparing the thickness of articular cartilage,which was significantly different from the FS PDWI sequence (P<0.05). Conclusion The 3D-FLASH sequence and derivated WE 3D-FLASH sequence have better definition of cartilage images and are mostly close to the real thickness,which possibly are the optimal scanning sequences for lumbar zypapophyseal joint articular cartilage MR imaging.更多还原