方法：前瞻性研究。选取我院2019-09/2020-12收治的干眼患者66例132眼，根据荧光素染色后泪膜破裂形态的不同分为片状破裂组(17例28眼)、类圆形破裂组(20例27眼)、线状破裂组(25例28眼)、点状破裂组(21例24眼)和不规则破裂组(20例25眼)，比较各组患者泪膜破裂动态变化图像特征、泪膜脂质层动态变化图像特征及泪膜首次破裂时间(NIBUTf)、泪膜平均破裂时间(NIBUTav)、泪河高度(TMH)、角膜荧光素染色(FL)评分的差异。

结果：各组患者NIBUTf有差异(P<0.001)，除点状破裂组和不规则破裂组间无差异(7.56±1.54s vs 8.02±1.86s，P=0.881)，其余各组两两比较均有差异(P<0.05)。各组间NIBUTav有差异(P<0.001)，除点状破裂组和不规则破裂组间无差异(9.54±2.12s vs 9.73±1.94s，P=0.997)，其余各组两两比较均有差异(P<0.05)。各组间TMH比较有差异(P<0.001)，除类圆形破裂组和线状破裂组间无差异(0.16±0.03mm vs 0.17±0.03mm，P=0.986)、点状破裂组和不规则破裂组间无差异(0.22±0.03mm vs 0.21±0.05mm，P=0.993)，其余各组两两比较均有差异(P<0.05)。各组患者FL评分和泪膜脂质层分级均有差异(P<0.001)。

结论：通过分析荧光素染色泪膜破裂的动态图像和脂质层扩散的动态图像特征，并结合其他泪膜静态检查参数，发现不同荧光素染色泪膜破裂形态可以直观反映干眼患者泪膜各层结构的变化，有助于临床医生识别干眼亚型，这对于干眼的诊断和分类具有潜在的临床价值。

METHODS: A prospective study. A total of 66 patients(132 eyes)with dry eye admitted to our hospital during September 2019 to December 2020 were divided area break-up group(17 cases, 28 eyes), circle break-up group(20 cases, 27 eyes), line break-up group(25 cases, 28 eyes), spot break-up group(21 cases, 24 eyes)and random break-up group(20 cases, 25 eyes)according to the different fluorescein tear film break-up patterns(BUPs). The image features of tear film break-up dynamic changes and the image features of TFLL dynamic changes were compared in each group, and the differences in the first non-invasive tear film break-up time(NIBUTf), average non-invasive tear film break-up time(NIBUTav), tear meniscus height(TMH)and fluorescein staining scores were compare,

RESULTS:A statistically significant differences were observed in NIBUTf among the groups of patients(P<0.001). Except that there was no difference between spot break-up group and random break-up group(7.56±1.54s vs 8.02±1.86s, P=0.881), other groups had significant differences in pairwise comparisons(P<0.05). There were statistically differences among the groups for NIBUTav(P<0.001). Except that there was no difference between spot break-up group and random break-up group(9.54±2.12s vs 9.73±1.94s, P=0.997), other groups had significant differences in pairwise comparisons(P<0.05). There was statistically difference among the groups for TMH(P<0.001). Except that there was no difference between circle break-up group and line break-up group(0.16±0.03mm vs 0.17±0.03mm, P=0.986), there was no difference between spot break-up group and random break-up group(0.22±0.03mm vs 0.21±0.05mm, P=0.993), other groups had significant differences in pairwise comparisons(P<0.05). There were statistically significant differences among the groups for FL scores and TFLL grading(P<0.001).

CONCLUSION:It is shown that different fluorescein BUPs intuitively reflects the tear film structure of the pathological changes according to evaluation and analysis of images feature of dynamic changes of FL tear film and TFLL combined with the results of static examination of tear film. It is helpful for clinicians to identify subtypes of dry eye, which has potential clinical value for the diagnosis and classification of dry eye.