重T2WI和增强T1WI MRI联合评估泪囊鼻泪管的优势

目的探讨MRI重T2WI(h-T2WI)和增强T1WI(Ce-T1WI)组合序列对正常和阻塞的泪囊鼻泪管的显示能力。材料与方法在脂肪抑制的基础上,用hT2WI和Ce-T1WI序列,薄层连续扫描正常和有阻塞的泪囊鼻泪管,扫描方位是轴面(AP)和冠状面(CP),用静态和动态两种扫描方式。结果静态扫描正常泪囊鼻泪管23例46侧,其中24侧用h-T2WI+Ce-T1WI+AP+CP组合,6侧用hT 2 W I+C e-T 1 W I+A P组合,8侧用h-T 2 W I+A P+C P组合,8侧用C eT1WI+AP+CP组合。动态轴面扫描正常泪囊鼻泪管10例20侧,均用hT2WI+AP组合。静态扫描阻塞的泪囊鼻泪管9例10侧,均用h-T2WI+CeT1WI+AP+CP组合。正常和有阻塞的泪囊鼻泪管均能被良好显示。(1)正常的泪囊鼻泪管:静态扫描见泪囊鼻泪管的管腔狭小,鼻泪管更小,并且形态多样;动态扫描见部分节段的管腔可自主性增大或变小。横断面上泪囊呈长椭圆形(16侧)或裂隙状(30侧),移行部均呈半月形,鼻泪管呈短椭圆形(28侧)或类圆形(18侧)。用静态h-T2WI序列,轴面图像上94.7%(36/38侧)的泪囊鼻泪管呈现了3层信号结构,冠状面图上31.2%(10/32侧)呈现了3层信号结构;这3层信号结构分别代表了管腔内容物、管壁内1/4和管壁外3/4;管腔内的泪液、泪膜和空气分别呈最高信号、高信号和最低信号;管壁内1/4呈低信号,管壁外3/4呈高信号。在Ce-T1WI序列上,管壁可以被明显强化。(2)有阻塞的泪囊鼻泪管:梗阻部位和病变的范围均被精确显示,其中管腔狭窄1侧,闭塞9侧;梗阻点以上管腔积液(脓)扩张,管壁变薄;梗阻点管腔消失或明显狭小,在h-T2WI上丧失3层信号结构。结论 h-T2WI结合Ce-T1WI序列的MRI,可以清晰的显示正常生理状态下的泪囊鼻泪管的管腔、管壁的层次和行经;也能够精确显示有梗阻的泪囊鼻泪管的梗阻部位、病灶范围,区分有血供和无血供的组织结构。动态h-T2WI发现正常管腔的大小有自主性变化。

Benefits of the combination of MRI heavily-T2WI and contrast-enhanced T1WI pulse sequences to examine human lacrimal sac and nasolacrimal duct

Objective: To evaluate capability of the combination of magnetic resonance imaging (MRI) heavily-T2 weighted imaging (h-T2WI) and contrast-enhanced T1 weighted imaging (Ce-T1WI) pulse sequences revealing the anatomic details of normal or obstructed human lacrimal sac (LS)-nasolacrimal duct (NLD). Materials and Methods:Using1.5TMRIsystem,thenormalandobstructedhuman LS-NLDs were imaged by h-T2WI and Ce-T1WI pulse sequences both with the fat-saturation technique. LS-NLD was scanned on its axial plane (AP) and coronal plane (CP). The thinnist slice thickness, consecutive sections (no spacing), static and dynamic scanning procedures were adopted. Results:Forty-six sides of normal LS-NLD (23 subjects) were imaged with the static scanning. Of them 24 sides were scanned with a combination of h-T2WI+Ce-T1WI+AP+CP, 6 sides with a combination of h-T2WI+Ce-T1WI+AP, 8 sides with a combination of h-T2WI+AP+CP, 8 sides with a combination of Ce-T1WI+AP+CP. The static scanning of 10 sides of obstructed LS-NLD (9 patients) was performed with a combination of h-T2WI+Ce-T1WI+AP+CP. The dynamic scanning of 20 sides of normal LS-NLD (10 subjects) was done with a combination of h-T2WI+AP. The normal and obstructed LS-NLDs were revealed vividly. (1)The normal LS-NLD. On the static scanning, their lumens were narrow and small, the size and shape of their lumens varied at different levels of LS-NLD, and NLD lumen was narrower and smaller than LS. On the dynamic scanning, the autonomous lumen changes (getting small or large) had been noticed. On cross sectional view, LS was long-ellipse (16 sides) or slit-shaped (30 sides), the junction was crescent, and NLD was short-ellipse (28 sides) or circular (18 sides). By h-T2WI, three-layer different signal intensities were revealed on a lot of images of LS-NLD, and 94.7%(36/38 sides) and 31.2%(10/32 sides) of LS-NLDs showed this signal pattern on axial and coronal scanning respectively. The three layer signals represented respectively (a) contents (tear, tear film or air) in their lumens, (b) medial part of their wall that was smaller than a quarter of their wall thickness and (c) lateral part of their wall that was more than three-fourths of their wall thickness. The tear, the tear film and the air in their lemen gave the most hyper-intense signal, middle-intense signal and the most hypo-intense signal respectively. The medial part and lateral part wall gave a hypo-and middle-intense signals respectively. The medial part wall consists of the epithelium layer, blood capillary layer and postcapillary venule layer, and the latter both are embedded in the lamina propria. The lateral part wall is venous lacunae layer embedded in the lamina propria, too. Ce-T1WI was able to enhance obviously the signal intensity of LS-NLD wall, therefor to promote distinguishing their wall from lumen (or contents). (2) The obstructed LS-NLD. Location and extent of the obstructive lesions were revealed precisely in all of the obstructed LS-NLDs. One side was lumen stricture and 9 sides were lumen occlusion. A lot of liquid (or pus) had been accumulated in the proximal LS-NLD lumen of the obstruction site, the lumen was dilatated, its wall was thinned and only showed one signal intensity. The lumen stricture might be revealed like an“hourglass”pattern. The site of lumen occlusion lost the normal three-layer signal pattern. Conclusions:MRI h-T2WI combined with Ce-T1WI pulse sequences can vividly reveal the lumen (or natures of the contents), wall and route of the normal human LS-NLD in normal physiological condition, and even can reveal detailed layers of their wall. They also can reveal the precise location and extent of lesions in the obstructed LS-NLDs. The tissues with blood supply can be distinguished from one (or the contents) without blood supply by this way. The autonomous size change of the normal LS-NLD lumens is revealed by the dynamic h-T2WI. This combination of MR pulse sequences can achieve aims of other imaging methods which are used to reveal LS-NLD and their lesions, so it is likely to replace the other approaches of dacryocystography with the combination.