磁共振检出肺小结节的序列优化及浸润性肺腺癌的诊断

发布时间：2018-05-16 13:48

本文选题：肺癌 + 肺结节　；参考：《河北医科大学》2017年硕士论文

【摘要】：第一部分磁共振检出小肺结节的序列优化目的:探讨磁共振扫描肺小结节的序列选择及其优化。方法:前瞻性收集肺部单发或多发≥6mm,≤20mm肺结节病人,于CT扫描完成24小时内行3-T磁共振扫描。扫描序列为T1-VIBE,T1-Star-VIBE,T2-TSE,HASTE序列,HASTE及T2-TSE序列分别各自增加240ms的反转恢复时间重新扫描。由两位放射学专家独立读片。计算或测量每个序列肺结节总检出率,实性结节检出率、磨玻璃结节检出率、CNR(对比噪声比)值及结节最大径。以CT的图像表现作为金标准,比较各序列结节总检出率、实性结节检出率、磨玻璃结节检出率、CNR及最大径是否有差异。采用SPSS21.0软件进行统计分析。计量资料的比较采用秩和检验,计数资料的比较采用χ2检验。P0.05认为差异有统计学意义。计算两位磁共振图像阅片者的一致性,统计学方法为Kappa检验。结果:2014年12月-2016年12月,共纳入肺结节病人136名(58.3±12.2岁),男性74名(59.8±10.1岁),女性62名(56.7±14.3岁)。共有结节145个。最大径为15.2±6.1mm。其中实性结节85例,最大径为14.7±6.3mm。磨玻璃密度结节(包括纯磨玻璃密度结节及混合密度结节)结节60例,最大径为14.2±5.2mm。结节总检出率为84.1%。其中实性结节共检出73例,总检出率为85.9%,实性结节在各序列的检出率分别为T1-VIBE(60.0%),STAR-VIBE(72.9%),HASTE(81.1%),HASTE-IR(82.4%),T2-TSE(76.5%),T2-TSE-IR(83.5%)。磨玻璃结节共检出49例,检出率为81.6%,磨玻璃结节在各序列的检出率分别为T1-VIBE(11.7%),STAR-VIBE(15.0%),HASTE(75.0%),HASTE-IR(78.3%),T2-TSE(66.7%),T2-TSE-IR(81.7%)。实性结节及磨玻璃结节的总检出率无统计学差异。各序列的CNR值分别为T1-VIBE(33.7±12.1),STAR-VIBE(95.1±33.2),HASTE(61.0±15.5),HASTE-IR(70.6±21.1),T2-TSE(47.3±12.2),T2-TSE-IR(71.9±22.1),有统计学差异。磁共振各序列显示结节的最大径相较于CT图像均偏小,且有统计学差异。两位阅片者一致性检验Kappa值为0.782(P0.001)。结论:3-T磁共振成像在肺结节检出的应用中有着巨大的潜力。HASTE-IR序列及T2-TSE-IR序列可用于≥6mm肺结节的检出。而且磁共振成像可用于磨玻璃结节的检出。第二部分磁共振浸润性肺腺癌的诊断目的:探讨磁共振在鉴别诊断以磨玻璃结节为表现的浸润性肺腺癌中的使用价值。方法:回顾性分析2014年12月-2016年12月在河北医科大学第四医院CT磁共振科行肺部CT及磁共振检查发现肺部磨玻璃结节患者。所有病人经外科手术切除且病理证实为癌前病变或临床分期为Ⅰ期早期肺腺癌。磁共振扫描序列为DWI,T1-Star-VIBE,T2-TSE,HASTE序列,HASTE及T2-TSE序列分别各自增加240ms的反转恢复时间重新扫描。由两位阅片者(阅片者1工作经验大于10年,阅片者2工作经验小于5年)在不知道病理结果的情况下分析病变的影像特征,有分歧者经两人商讨后达成一致。测量磁共振图像以下指标:(1)病变大小;(2)T2信号强度;(3)结节ADC值。依据病理结果进行分组,将浸润前病变(包括AAH、AIS)与微浸润性病变归为一组,即组A;浸润性病变为一组,即组B。比较两组病人的3种指标是否有差异。采用SPSS21.0软件进行统计分析。计量资料如满足正态分布采用两个独立样本t检验,若不满足正态分布采用Mann-Whitney U检验计。计数资料使用卡方检验。使用ROC曲线进行诊断实验评价。P0.05认为差异有统计学意义。结果:最后纳入人群共共34人,年龄55.8±12.0岁。其中男性14人,年龄57.1±14.7岁;女性20人,年龄54.9±10.1岁。共纳入结节34例。组A 15例,包括AAH4例,AIS2例,MIA9个,组B 19例。组B中女性比例高于组A中女性患者的比例(69.4%vs 46.7%),但无统计学差异(χ2=0.1.638,P=0.201)。组B中患者发病年龄小于组A中患者发病年龄(54.3±13.1岁vs 57.6±10.8岁),但无统计学差异(t=0.786,P=0.438)。组A中病变直径小于组B中病变直径(9.9±2.6mm vs 13.1±2.7mm),且有统计学差异(t=-3.405,P=0.002)。组A中病变的T2信号强度低于组B中的病变直径(93.0±8.3 vs 113.6±22.9),且统计学差异(t=-3.6,P=0.001)。组A中病变ADC值要低于组B(1.0±0.2*10-3 mm2/s vs 1.3±0.3*10-3 mm2/s),且有统计学差异(t=-2.697,P=0.011)。ROC曲线分析,区分组A病变与组B病变的最佳指标是病变大小,界值为11.5mm,敏感度为73.7%,特异度为73.3%。AUC值为0.791(95%可信区间:0.640,0.942)。结论:3-T磁共振,使用结节大小、T2信号强度、ADC值,对于鉴别诊断浸润性肺腺癌是有帮助的。磁共振鉴别浸润性肺腺癌与浸润前病变(包括MIA)的最佳指标是结节大小。
[Abstract]:Partial optimization of the sequence of small pulmonary nodules detected by magnetic resonance: To explore the sequence selection and optimization of magnetic resonance scanning pulmonary nodules. Methods: a prospective collection of pulmonary nodules with single or more 6mm or less 20mm pulmonary nodules was prospectively collected for 24 hours by CT scan. The sequence was T1-VIBE, T1-Star-VIBE, T2-TSE, HASTE sequence. The HASTE and T2-TSE sequences respectively increased the reversal time of 240ms, respectively. The total detection rate of pulmonary nodules, the detection rate of solid nodules, the detection rate of ground glass nodules, the value of CNR (contrast noise ratio) and the maximum diameter of nodules were calculated or measured by two radiologists. The image performance of CT as the gold standard was compared with each other. The total detection rate of serial nodules, the detection rate of solid nodules, the detection rate of glass nodules, CNR and the maximum diameter were different. The statistical analysis was carried out by SPSS21.0 software. The comparison of the data was compared with the rank sum test, and the comparison of the counting data using the x 2 test.P0.05 thought the difference was statistically significant. Results: Kappa test. Results: in December December 2014 -2016, 136 patients with pulmonary nodules (58.3 + 12.2 years old) were included, male 74 (59.8 + 10.1 years), female 62 (56.7 + 14.3 years), 145 nodules with a maximum diameter of 15.2 + 6.1mm., 85 cases of solid nodules, and maximum diameter of 14.7 + 6.3mm. mill glass density nodules (including pure grinding glass). 60 cases of nodules and mixed density nodules were found in 60 cases. The total detection rate of the maximum diameter of 14.2 + 5.2mm. nodules was 84.1%. and the total detection rate was 85.9%. The detection rates of solid nodules were T1-VIBE (60%), STAR-VIBE (72.9%), HASTE (81.1%), HASTE-IR (82.4%), T2-TSE (76.5%), T2-TSE-IR (83.5%), and T2-TSE-IR (83.5%). 49 cases of glass nodules were detected, the detection rate was 81.6%, the detection rates of the glass nodules were T1-VIBE (11.7%), STAR-VIBE (15%), HASTE (75%), HASTE-IR (78.3%), T2-TSE (66.7%), T2-TSE-IR (81.7%). The total detection rates of solid nodules and grinding glass nodules were not statistically different. The CNR values of each sequence were T1-VIBE (33.7 + 12.1), STAR-V, respectively. IBE (95.1 + 33.2), HASTE (61 + 15.5), HASTE-IR (70.6 + 21.1), T2-TSE (47.3 + 12.2), T2-TSE-IR (71.9 + 22.1), there were statistically significant differences. The maximum diameter of the nodules was smaller than that of CT images, and there were statistical differences. The Kappa value of the two film conformance test was 0.782 (P0.001). Conclusion: 3-T magnetic resonance imaging is in the pulmonary nodule. There are huge potential.HASTE-IR sequences and T2-TSE-IR sequences in the detection of pulmonary nodules over 6mm. And magnetic resonance imaging can be used for the detection of glass nodules. The second part of the MRI diagnosis of invasive lung adenocarcinoma: the study of magnetic resonance in the differential diagnosis of invasive lung adenocarcinoma with glass nodule as the manifestation Methods: a retrospective analysis was made to a retrospective analysis of pulmonary CT and magnetic resonance imaging in the CT MRI Department of the fourth hospital of Hebei Medical University, December 2014, in December -2016. All patients were excised by surgical excision and were confirmed by pathology as precancerous lesions or clinical stages as early stage of lung adenocarcinoma. Magnetic resonance scanning sequence. DWI, T1-Star-VIBE, T2-TSE, HASTE sequences, HASTE and T2-TSE sequences respectively increased the reversal time of 240ms, respectively. The image features of the lesions were analyzed under the condition of not knowing the pathological results without the pathological results, and the disagreement was discussed by two people. The following indexes of MRI images were measured: (1) lesion size; (2) T2 signal intensity; (3) ADC value of nodules. According to the pathological results, the preinvasive lesions (including AAH, AIS) and microinvasive lesions were classified as a group, namely, group A; the group of invasive lesions was a group, that is, the 3 indexes of group B. compared to the two groups were different. SPSS21.0 soft. Two independent sample t tests were used to satisfy normal distribution. If the normal distribution was not satisfied with the normal distribution, the Mann-Whitney U test was used. The counting data was checked by chi square test. The diagnostic test of the ROC curve was used to evaluate.P0.05. The difference was statistically significant. Finally, 34 people were included in the population, and the age was 55.8 + 12. .0 years old, of which 14 men, age 57.1 + 14.7 years, women 20, age 54.9 + 10.1 years old, were included in 34 cases. Group A 15 cases, including AAH4, AIS2, MIA9, and B 19. The proportion of women in group B was higher than that of women in A (69.4%vs 46.7%), but there was no statistical difference (chi 2=0.1.638, P=0.201). The onset age of the patients (54.3 + 13.1 years vs 57.6 + 10.8 years old), but there was no statistical difference (t=0.786, P=0.438). The diameter of the lesion in group A was less than that in group B (9.9 + 2.6mm vs 13.1 + 2.7mm), and there were statistical differences (t=-3.405, P=0.002). The intensity of the T2 signal in group A was lower than that in the group (93 + 8.3 113.6 + 22.9), and statistics The difference (t=-3.6, P=0.001). The ADC value in group A was lower than that of group B (1 + 0.2*10-3 mm2/s vs 1.3 + 0.3*10-3 mm2/s), and there was a statistical difference (t=-2.697, P=0.011) curve analysis. The best index to distinguish between group and group was the size of lesion, the boundary value was 73.7%, and the specificity was 0.791 (95% confidence interval: 0.640,0.942) conclusion: 3-T magnetic resonance, using nodule size, T2 signal intensity, and ADC value, is helpful in the differential diagnosis of invasive lung adenocarcinoma. The best indicator of the differential diagnosis of invasive lung adenocarcinoma and preinvasive lesions (including MIA) is the size of the nodules.

【学位授予单位】：河北医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R445.2;R734.2

【参考文献】