CT 肺灌注在肺结节诊断中的应用研究

CT 肺灌注在肺结节诊断中的应用研究
张金娥 梁长虹 赵振军 林华欢 曾辉 何晖 茹光腾
【摘要】 目的 研究CT 灌注成像对良、恶性肺结节的诊断价值。方法 前瞻性研究88 例直
径2 ～4 cm 的肺结节的多层螺旋CT 灌注表现。其中肺癌62 例, 良性病变26 例( 炎性假瘤12 例, 结
核球10 例, 错构瘤3 例, 曲菌球1 例) 。采用8 层螺旋CT 灌注成像, 电影模式, 层厚5 mm, 4 层/ 圈, 扫
描时间1s / r, 数据采集时间40 s。碘普胺( 300 mg I /ml) 50 ml, 用高压注射器经前臂浅静脉注射, 流率
4 ml / s, 延迟5. 6 s。CT Perfusion 2 软件分析测量结节的血流量( BF) 、血容量( BV) 、平均通过时间
( MTT) 、表面通透性( PS) 和拟合时间-密度曲线。结果 良、恶性结节的BV 值( 分别为5. 33、
10. 00 ml /100 g) 和PS 值( 分别为13. 11、44. 94 ml·100 g - 1 ·min - 1 ) 差异有统计学意义( F 值分别为
29. 368 和48. 027, P 值均为0. 000) 。以BV≥6 ml /100 g 作为恶性病变的阈值, 其敏感度87. 3%, 特
异度100. 0% , 阳性预测值100. 0% , 阴性预测值65. 4% , 准确度89. 9% 。以PS≥30 ml ·100 g - 1 ·
min - 1 作为恶性结节的阈值, 则敏感度96. 4% , 特异度75. 0% , 阳性预测值87. 1%, 阴性预测值
87. 2% , 准确度88. 6%。以BV≥6 ml / 100 g 同时PS≥30 ml·100 g- 1 ·min - 1 作为恶性结节的阈值,
则敏感度96. 8% , 特异度92. 3% , 阳性预测值96. 8%, 阴性预测值92. 3% , 准确度95. 5% 。肺癌的拟
合时间-密度曲线和主动脉的关系密切, 起始端与主动脉同步, 峰值在主动脉峰值区域或稍后, 85. 5%
( B 型+ C 型, 53 / 62) 呈缓慢下降或平台型。结论 多层螺旋CT 灌注对肺结节的良恶性鉴别诊断有
较大帮助。
【关键词】 硬币病变, 肺; 体层摄影术, X 线计算机; 灌流
Ut ility of CT perfusion in pulmona ry nodules ZHANG Jin-e, LIANG Chang-hong, ZHAO Zhen-jun,
LIN Hua -huan, ZENG Hui, HE Hui, RU Guang-teng. Depa rtment of Ra diology, Gua ngdong Provincia l
People′s Hospital, Guangzhou 510080, China
【Abstr act 】 Objective To evaluate the diagnosis value of CT perfusion in pulmonary nodules.
Methods A prospective study was undertaken in eighty-eight patients with the diameter of 2 - 4 cm of
pulmonary nodules using MSCT perfusion. They were proved with pathology as peripheral lung cancer in
62 cases, benign nodules in 26 cases( 12 cases of inflammatory pseudotumor, 10 cases of tubercle nodule,
3 cases of hamartoma, 1 case of aspergilloma) . Using 8-row detector spiral CT, cine scan, slice thickness
5 mm/4 slices, scan time was 1 second per circle, acquisition time was 40 seconds. Ultravist( 300 mg I /ml)
was administered at a delay of 5. 6 seconds rate of 4 ml / s from forearm superficial for a total of 50 ml and a
vein by hyperbaric injection. The CT data were transferred to a workstation ( Sun Microsystems, Advantage
Windows 4. 0, GE Medical Systems ) and analyzed by a CT Perfusion 2-body tumor software. The timedensity
curve and parameters which include blood flow( BF) , blood volume( BV) , mean transit time( MTT)
and permeability ( PS) of CT perfusion were analyzed. Results The differences of BV and PS were
statistically significant between lung cancer to benign nodules ( F = 29. 368, P = 0. 000 and F = 48. 027, P =
0. 000) . When BV≥6 ml /100 g was set as the diagnostic threshold, the sensitivity, specificity, positive
predict value, negative predict value and accuracy were 87. 3% , 100. 0% , 100. 0% , and 65. 4% ,
respectively. When PS≥ 30 ml · 100 g - 1 · min - 1 was set as the diagnostic threshold, the sensitivity,
specificity, positive predict value, negative predict value and accuracy were 96. 4% , 75. 0% , 87. 1% ,
87. 2% , and 88. 6%, respectively. When BV≥6 ml /100 g and PS≥30 ml·100 g - 1 ·min - 1 were set as
the diagnostic threshold, the sensitivity, specificity, positive predict value, negative predict value and
accuracy were 96. 8% , 92. 3% , 96. 8%, 92. 3% , and 95. 5% , respectively. The time-density curve of
lung cancer was correlation to that of aorta. The beginning of the curve was synchronization and the location
of the peak was same or slightly delayed in lung cancer to aorta. The time-density′s form ( type B + type C)
of lung cancer ( 85. 5% ) was descent slowly or flat. Conclusion MSCT perfusion is helpful to the
diagnosis and differential diagnosis of aorta pulmonary nodules.
【Key wor ds】 Coin lesion, pulmonary; Tomography, X-ray computed; Perfusion

CT 肺灌注在肺结节诊断中的应用研究胸部.rar