传统核探测器故障信号诊断研究都需要提前提取信号特征,然后用机器学习、支持向量机、统计方法等对特征进行分类.为了实现对探测器输出信号进行实时识别和故障诊断,本文基于Matlab平台构建了一个用于对图像进行分类的卷积神经网络模型,对核探测器故障信号进行分类诊断.从分类准确率和算法运行时间两个方面对Adam、Sgdm、Rmsprop三种优化算法进行了比较.结果表明Rmsprop算法运行时间最少,但准确度和损失的训练迭代曲线不平稳;Sgdm模型对十组非正常信号图像分类的准确率最高为93.10％,准确度和损失的训练迭代曲线平稳.虽然,本文方法诊断准确率略低于文献报道值,但是不需要对信号进行预处理和特征预提取,使用更为简便.

Cathode current is a key measure of the rated power of the diode and is closely related to the cathode structure. Compared with $\Pi $ cathode and helical cathode, the mesh cathode has been more widely used in high-power tetrode, because of its structural advantage. So far, the relationship between the space charge limited current (SCLC) and the cathode structure for thermionic emission mesh cathode diode has not been reported. In this article, we develop a calculation expression of the SCLC for the mesh cathode coaxial cylindrical diode, which is based on the calculation model of the helical cathode. The results show that the calculated values are in excellent agreement with the simulated values based on CST Particle Studio. In addition, we study the dependence of the SCLC of the mesh cathode diode with the change of the cathode height, radius of the cathode filament, anode voltage, and density of the mesh. It is helpful for the optimization of the cathode structure design of vacuum tubes and microwave tubes with the coaxial cylinder mesh cathode structure.

As artificial intelligence technology has progressed, numerous businesses have used intelligent diagnostic technology. This study developed a deep LSTM neural network for a nuclear power plant to defect diagnostics. PCTRAN is used to accomplish data extraction for distinct faults and varied fault degrees of the PCTRAN code, and some essential nuclear parameters are chosen as feature quantities. The training, validation, and test sets are collected using random sampling at a ratio of 7:1:2, and the proper hyperparameters are selected to construct the deep LSTM neural network. The test findings indicate that the fault identification rate of the nuclear power plant fault diagnostic model based on a deep LSTM neural network is more than 99 percent, first validating the applicability of a deep LSTM neural network for a nuclear power plant fault-diagnosis model.

The helical cathode is one of the classical structures of the cathode for the microwave tube, and has a wide application in pulse transmission tube. In this article, we propose a quantitative expression of the space charge limited current (SCLC) of the helical cathode diode through improve the theoretical model of Japanese distinguished scholar Y. Koike, and study the dependence of the SCLC of the helical cathode diode with radius of the cathode filament ( $r_{f}$ ), cathode height ( $H$ ), helical pitch ( $d$ ), and anode voltage ( $V$ ), respectively. The results show that the value of current is an increasing function of $r_{f}$ , $H$ , and $V$ , and a decreasing function of $d$ . The simulation results show that when the radius of the cathode filament is greater than or equal to 1.6 mm, the proposed model of this article is more suitable than Y. Koike’s model, and can be used in combination with Y. Koike’s model to estimate quantitatively the SCLC of helical cathode diode. Moreover, it also provides an effective basis for the design and optimization of the helical cathode structure of the microwave tube.

Objective:Topredict the three-dimensional dose distribution of regions of interest (ROI) with brachytherapy for cervical cancer based on U-Net fully convolutional network, and evaluate the accuracy of prediction model.Methods:First, 100 cases of cervical cancer intracavity combined with interstitial implantation were selected as the entire research data set, and divided into the training set ( n=72), validation set ( n=8), and test set ( n=20). Then the U-Net was used to construct two models based on whether the uterine tandem and the implantation needles were included as the distinguishing factors. Finally, dose distribution of 20 cases in the test set were predicted using the trained model, and comparative analysis was performed. The performance of the model was jointly evaluated by , and the mean absolute deviation (MAD). Results:Compared with the model without the uterine tandem and the implantation needles, the of the rectum was increased by (16.83±1.82) cGy ( P<0.05), and the or of the other ROI were not different significantly (all P>0.05). The MAD of the high-risk clinical target volume, rectum, sigmoid, small bowel, and bladder was increased by (11.96±3.78) cGy, (11.43±0.54) cGy, (24.08±1.65) cGy, (17.04±7.17) cGy and (9.52±4.35) cGy, respectively (all P<0.05). The MAD of the intermediate-risk clinical target volume was decreased by (120.85±29.78) cGy ( P<0.05). The mean value of MAD for all ROI was decreased by (7.8±53) cGy ( P<0.05), which was closer to the actual plan. Conclusions:U-Net fully convolutional network can be used to predict three-dimensional dose distribution of patients with cervical cancer undergoing brachytherapy. Combining the uterine tube with the implantation needles as the input parameters yields more accurate predictions than a single use of the ROI structure as the input.

Developing a fluorescent probe for UO22+, which is resistant to interference from other ions such as Cu2+ and can be applied in acidic and high-water systems, has been a major challenge. In this study, a “turn-off” fluorescent probe for triamine-modified flavonoid derivatives, 2-triphenylamine-3-hydroxy-4H-chromen-4-one (abbreviated to HTPAF), was synthesized. In the solvent system of dimethyl sulfoxide:H2O (abbreviated to DMSO:H2O) (v/v = 5:95 pH = 4.5), the HTPAF solution was excited with 364 nm light and showed a strong fluorescence emission peak at 474 nm with a Stokes shift of 110 nm. After the addition of UO22+, the fluorescence at 474 nm was quenched. More importantly, there was no interference in the presence of metal ions (Pb2+, Cd2+, Cr3+, Fe3+, Co2+, Th4+, La3+, etc.), especially Cu2+ and Al3+. It is worth noting that the theoretical model for the binding of UO22+ to HTPAF was derived by more detailed density functional theory (DFT) calculations in this study, while the coordination mode was further verified using HRMS, FT-IR and 1HNMR, demonstrating a coordination ratio of 1:2. In addition, the corresponding photo-induced electron transfer (PET) fluorescence quenching mechanism was also proposed.

本文基于同轴二极管一维空间电荷限制流经典理论公式,利用CST粒子工作室研究了热电子向外发射同轴二极管二维空间电荷限制流随二极管几何结构参数变化的规律和几何结构对虚阴极产生临界条件的影响.结果表明,二极管修正系数是纵横比的单调递减函数;随阳极半径与阴极半径比值的增大,修正系数逐渐降低;二极管的电压、阳极和阴极半径及阴极长度均会影响虚阴极的产生.本文研究结果将为射频四极管结构设计及优化提供参考和理论依据.

目的:比较左侧乳腺癌全切术后无均整器(FFF)与均整器(FF)模式下容积旋转调强放疗(VMAT)计划剂量学特点,分析无均整器模式下剂量学优势.方法:随机选取2019～2020年左侧乳腺癌全切术后的患者10例,分别用两种模式对同一患者的计划靶区(PTV)给予相同的处方剂量50 Gy,在RayStation计划系统上设计FFF和FF的VMAT计划,在满足临床要求的情况下,比较两组剂量学特点、机器总跳数、出束总时间.结果:FFF模式靶区PTV的D98％低于FF模式(P=0.007),D95％、D2％、CI、HI两者比较差异均无统计学意义(P>0.05);FFF模式在患侧肺、全肺、心脏、脊髓以及冠状动脉前降支保护上明显优于FF模式(P<0.05),患侧肺平均剂量同比降低9.59％,心脏平均剂量降低12.20％;在健侧肺、臂丛以及正常组织的保护上两者比较差异无统计学意义(P>0.05);FFF模式的机器总跳数相对于FF模式增加1.30倍,出束总时间缩短至FF模式的87.15％.结论:在两组治疗计划均能满足临床治疗要求的前提下,FFF模式可显著减少患者危及器官以及正常组织的受照剂量,虽然机器总跳数增加,但患者的治疗时间反而下降.因此可知FFF模式优于FF模式的治疗效率.

目的:量化分析后装治疗中,使用不同累积剂量计算方法导致的总剂量评估差异.方法:使用混合形变配准算法,对32例已完成后装插植治疗的宫颈癌患者共计进行108次形变配准,并计算每个患者的形变总剂量.根据膀胱和直肠的体积控制情况将32例患者分为体积控制组(18例)和体积未控制组(14例),计算比较组间累积剂量体积参数(D0.1 cc、D1 cc、D2 cc)和相似度系数(DSC),并与GEC-ESTRO推荐的分次间剂量累加方法相比较.结果:DSC随形变图像相对体积比值的增大而变差.体积控制组中,GEC-ESTRO推荐方法和形变累积剂量差异(DDIR/DGEC)较小.对于膀胱,体积控制组D2 cc的DDIR/DGEC最大值为1.05,低于未控组1.07;DDIR/DGEC最小值为0.91,高于未控组0.80.提示在使用形变剂量评估过程中,形变剂量值可能比GEC-ESTRO推荐的剂量计算方法偏高5％～7％或偏低9％～20％,偏高和偏低的原因主要为参与形变的体积差异导致剂量网格缩放.而D0.1 cc、D1 cc的DDIR/DGEC更接近1.对于直肠,控制组没有观察到DDIR/DGEC偏高的情况,D2 cc的DDIR/DGEC最大值为0.99,低于未控组1.06;最小值为0.80,低于未控组0.85,但未控制组平均偏高6％.而直肠的D0.1 cc、D1 cc、D2 cc的累积剂量比值都较接近,并且差异不如膀胱显著.结论:在后装治疗中,形变体积可以影响到形变累积剂量的计算结果,合理控制形变轮廓的体积不仅有助于提高形变的准确性,而且形变总剂量评估结果更稳定,建议在后装治疗过程中重视对膀胱和直肠的体积控制.

The anode and the cathode of the cylindrical $\Pi $ cathode diode are not coaxial, and the quantitative calculation of its space charge limited current (SCLC) has not been reported. In this article, we propose the SCLC quantitative expression of cylindrical $\Pi $ cathode diode through improving the theory model of Japanese distinguished scholar Y. Koike ( ) based on the 3/2 power law of SCLC of the coaxial cylindrical diode and study the relationship between current and cathode structure of cylindrical $\Pi $ cathode diode by using the numerical simulation method based on the CST Particle Studio (CST-PS). This improved model is more suitable than previous empirical formulas when the radius of a single cathode is greater than or equal to 1.0 mm and can be used in combination with the theory model of Y. Koike to estimate quantitatively the SCLC of cylindrical $\Pi $ cathode diode. Moreover, it also provides an effective basis for the design and optimization of the $\Pi $ cathode structure of the high-power vacuum transmitting tube.

国内核燃料元件公司以前采用的抽插力测量系统存在设备陈旧,机械故障多,系统不能稳定运行等不足,通过对燃料组件抽插力测量系统进行研究,设计了新的机械结构和数字化系统的改进方案,并在燃料元件厂实际使用中实现了燃料组件抽插力测量的数据自动读取与测量底座自动转换.研究结果表明,新的燃料组件抽插力测量系统实现了燃料组件抽插力测量过程自动化、数字化和系统化的提升,系统更加稳定,生产噪声明显变小,操作更安全便捷.

Objective To compare the dosimetric differences between free-hand method and virtually optimized method for implanting needles in intracavitary and interstitial combined brachytherapy (IC/IS BT) of cervical cancer,and to explore the improvement space of the existing interstitial brahcytherapy plan.Methods High-dose-rate cervical cancer IC/IS BT plans (short for Treatment-Plan) of 18 cases were retrospectively analyzed.For each treatment plan,Nucletron Oncentra 3D brachytherapy planning system was utilized to redesign the virtually optimized insertion method IC/IS BT plan (short for Optimized-Plan).Dose volume histogram was adopted to evaluate the dose distribution in high-risk clinical target areas and exposure dose to organ at risk (OAR).The plan execution efficiency between two plans was also assessed.Results Comparing these two plans,the differences in conformity and uniformity of dose distribution of the target area were statistically significant (P=0.000,0.008).The differences of D0.01 cm3,D1 cm3,D2 cm3 and D5 cm3 in bladder,rectum,sigmoid and small bowel were all statistically significant (all P＜0.05).Optimized-Plan could reduce the D2 cm3 of bladder,rectum,sigmoid and small bowel by 60.41,36.43,27.53 and 12.43 cGy,respectively.The execution time for the Treatment-Plan and Optimized-Plan were (857.92±243.39) s and (804.53±239.13) s with statistical significance (P＜0.001).Conclusions Compared with the free-hand method,virtually optimized method yields more conformable coverage of the target area and more uniform dose distribution.At the same time,the doses of each OAR are reduced to different degrees and the execution time of the plan is also shortened.

通过使用软件Varian公司的Trilogy加速器自带的EPID配合EDose软件和德国IBA公司的RAROR Chamber以及与其配套的DOSE1剂量仪和RW-3固体水分别对12例锥筒计划进行绝对剂量验证,记录并分析使用EDose软件时在不同标准下绝对剂量γ通过率和使用电离室时的点剂量验证结果.发现采用γ(3％/3 mm)、(3％/2 mm)、(2％/2 mm)标准时,EDose软件验证的绝对剂量 γ 通过率分别为(99.59％±0.33％)、(96.96％±1.27％)、(95.48％±1.37％),等中心处绝对剂量偏差为0.27％±1.02％;电离室测量的等中心处绝对剂量偏差为-0.03％±1.20％,最大偏差是-1.82％.两种验证方法的验证结果均满足临床治疗的要求且无统计学差异.结果表明基于电子射野影像系统(EPID)的EDose软件可以用于锥筒计划的剂量验证,而且可以提供二维剂量分布,操作更加方便,同时可以节约科室成本.

核探测器是一种特殊的随机信号转换器,发生故障时,采用传统的人工方法很难及时有效地对故障进行诊断.本文提出了一种针对闪烁体探测器信号波形特征的在线智能故障检测与分类方法,通过分析闪烁体探测器不同故障时的输出信号变化特征,建立了相应的故障模型.使用小波包算法与支持向量机理论分解并提取特征向量,即可判断故障类型.以ST401闪烁体探测器为例,进行了模拟仿真实验.实验结果表明,基于信号波形特征的数字化方法能快速有效地对闪烁体探测器进行故障自动诊断.

目的:定量分析剂量网格分辨率的大小对非小细胞肺癌(NSCLC)立体定向放射治疗(SBRT)计划剂量分布的影响,指导临床选用合适的剂量网格分辨率用于肺癌SBRT计划设计.方法:选取10例NSCLC患者,采用容积旋转调强技术,使用0.20 cm的剂量网格分辨率设计SBRT计划,将治疗计划结果再分别用0.40、0.30、0.25、0.15、0.10 cm的剂量网格分辨率计算最终剂量.比较6种不同剂量网格分辨率下的计划靶区(PTV):D2％、Dmean、D98％、均匀性指数(HI)、适形度指数(CI)和危及器官:全肺、胸壁、食管、心脏、脊髓、主动脉、气管树等的相关剂量学参数的差异.结果:与0.20 cm剂量网格分辨率组计划相比较,0.40、0.30、0.25、0.15、0.10 cm剂量网格分辨率计算得到的计划的靶区D2％、Dmean、CI均具有统计学意义(P<0.05);HI除0.15 cm剂量网格分辨率组以外,均具有统计学意义(P<0.05).在危及器官受照剂量方面,与0.20 cm剂量网格分辨率组计划相比较,大于0.20 cm(0.40、0.30、0.25 cm)组计划全肺、胸壁、食管、心脏、脊髓、主动脉、气管树等的相关剂量学参数差异均有统计学意义(P<0.05);小于0.20 cm(0.15、0.10 cm)组计划除全肺V10、V20、V12.5和V13.5有影响以外(P<0.05),胸壁、食管、心脏、脊髓、主动脉、气管树等的相关剂量学参数差异较小(P>0.05).结论:剂量网格分辨率的大小会影响剂量计算的准确性,在NSCLC患者SBRT计划设计时,建议使用0.20 cm或更小的剂量网格分辨率.

目的 初步探讨固定剂量率容积调强在直肠癌术后盆腔放射治疗中的可行性.方法 选取10例直肠癌放疗患者,利用RayStation计划系统为每例患者制定可变剂量率容积调强(VDR-VMAT)、固定剂量率容积调强(CDR-VMAT)和5野静态调强(5F-sIMRT)计划.运用剂量体积直方图评估三种计划的靶区、危及器官和正常组织剂量学参数,评估机器总跳数(MU)和计划执行时间.结果 靶区剂量方面,三种计划的靶区D2％、Dmean、D98％、HI和CI在总体上均有差异(P<0.001).CDR-VMAT与VDR-VMAT相比较,D2％、Dmean、D98％、HI、CI差异均无统计学意义;CDR-VMAT与5F-sIM-RT相比较,D2％减小1.55 Gy(P=0.005)、Dmean减小0.99 Gy(P=0.005)、D98％增大0.60 Gy(P=0.03)、HI值减小(P=0.008)、CI值增大(P=0.008).危及器官方面,三种计划的膀胱Dmean、V45、V40,小肠Dmean、Dmax、V45,左右股骨头Dmean在总体上均有差异(P<0.05);小肠V40、V35,左右股骨头V45、V40在总体上均无差异(P>0.05).CDR-VMAT与VDR-VMAT相比较,膀胱、小肠、左右股骨头的各剂量学参数差异无统计学意义;CDR-VMAT与5F-sIMRT相比较,膀胱Dmean减小3.05 Gy(P=0.005)、V40减小0.88％(P=0.042),小肠Dmean减小1.75 Gy(P=0.002)、Dmax减小1.70 Gy(P<0.001),左、右股骨头Dmean减小(P=0.008,0.042).正常组织低剂量受照体积方面,三种计划的正常组织受照低剂量体积除V10(P=0.497)之外,V5、V15、V20、V25和V30在总体上均有差异(P<0.001).CDR-VMAT与VDR-VMAT相比,V5、V10、V15、V20、V25和V30差异无统计学意义;与5F-sIMRT相比,CDR-VMAT的V5减小1.18％(P=0.005)、V15减小0.61％(P=0.022)、V30减小0.80％(P=0.022),V10、V20和V25差异无统计学意义.CDR-VMAT计划的MU为(668.51±45.92),比VDR-VMAT(574.13±50.20)增加16.44％,比5F-sIMRT(537.19±37.34)增加24.45％;CDR-VMAT计划执行时间(3.34±0.22)min是VDR-VMAT(1.76±0.04)min的近两倍,比5F-sIMRT(4～6)min稍短.结论 CDR-VMAT可形成与VDR-VMAT一样高质量的计划,较5F-sIMRT有更优的靶区覆盖率、危及器官保护和正常组织低剂量受照体积.但CDR-VMAT计划的MU比VDR-VMAT和5F-sIMRT增多;执行时间比VDR-VMAT长,比5F-sIMRT稍短.CDR-VMAT初始成本较低,有望为不具备可变剂量率的直线加速器提供额外的旋转放疗的选择.

目的 基于OSL原理,测量和分析在激发光下剂量片中沉积的剂量与激发出来的荧光光子数目的关系,为以后相关科研提供一个可靠的依据.方法 使用一个1.48 GBq(40 mCi)-90Sr/90Y β源对α-Al2O3∶C剂量片进行不同时间的辐照,再用设计好的激发光装置将沉积在α-Al2O3∶C剂量片的信号进行光激发,通过光电倍增管对荧光光子进行收集,最后通过单光子计数器输出计数.结果 通过上面的实验方法得到了剂量片上沉积剂量与荧光光子数的关系,通过线性拟合后可以得到线性关系.结论 从结果上可以看出剂量片上的本底剂量与其产生的荧光光子数存在线性关系,同时沉积剂量与荧光光子数也呈线性关系.

Objective To evaluate the dosimetric and biological advantages of equivalent uniform dose optimization based on organs-at-risk (OAR) partition constraint in volumetric modulated arc therapy (VMAT) for rectal cancer. Methods With dose-volume constraints, 10 VMAT plans for rectal cancer which were approved in clinic were selected (PHY plans). Based on the PHY plans, we performed the OAR partition according to the position relationships among the small intestine, the bladder and target areas, and carried out the equivalent uniform dose (EUD) optimization by giving different specific-tissues parameters (a value) which were different in sensitivity to hot spots, finally obtaining EUD plans. The greater a value was more sensitive to hot spots, and the dose of overlaps between target areas and OAR was relatively higher than that of other regions, with a higher probability of occurrence of hot spots, thus we selected a larger a value (a=10) to perform EUD optimization. The area where 9 mm expanded from the overlapping portions was intersected with the small intestine and the bladder had a low incidence of hot spots, therefore a value was at 8 and 5. As the rest of small intestine and bladder were away from high-dose area, we took a value of 6.7 and 2.3. The dosimetric and biological features of EUD plans and PHY plans were compared and SPSS 22 software was used to analyze the obtained data. Results The differences between PHY plan and EUD plan in PTV parameters were trivial. However, compared with PHY plans, EUD plans had reduced OAR dose parameters and normal tissue complication probability, with statistically significant differences (P<0.05). In EUD plan, the mean dose of the small intestine and the bladder decreased by 4.5 and 6.8 Gy, respectively;NTCP decreased 40.0％and 6.6％, respectively. Conclusion EUD optimization based on OAR partition constraint in VMAT for rectal cancer obviously decreases the OAR irradiation dose and normal tissue complication probability, without affecting PTV, which is significant to improve the quality of the plan.

Objective To compare the biological and dosimetric differences of volumetric modulated arc therapy (RapidArc) and fixed-field intensity-modulated radiotherapy (IMRT) based on generalized equivalent uniform dose (gEUD) optimization for rectal cancer and provide some reference for clinical decision. Methods Ten patients with rectal cancer were selected, and three kinds of plans were designed for each patient, namely five-field IMRT (5FIMRT), seven-field IMRT (7FIMRT) and RapidArc plans, with the same prescription dose of 50 Gy/25 F for planning target volume (PTV). Finally, we applied paired t test to compare the differences among three plans in the mean dose-volume histogram and monitor unit, the conformity index (CI) and homogeneity index (HI) of PTV, the normal tissue complication possibility of organs-at-risk, the gEUD and physical dose of PTV and organs-at-risk. Results All the target dose distribution in three plans met the clinical requirement. The CI and HI of RapidArc were better than those of 5FIMRT and 7FIMRT (P<0.05). In the protection of small intestine and bladder, 7FIMRT and RapidArc had no significant differences (P>0.05), and both of the two plans are better than 5FIMRT (P<0.05). For the protection of bilateral femoral head, RapidArc was superior to 7FIMRT (P<0.05), and 7FIMRT was better than 5FIMRT (P<0.05). RapidArc had significantly less monitor units as comparison with 5FIMRT and 7FIMRT (P<0.05). Conclusion In the plans based on gEUD optimization, RapidArc and 7FIMRT were superior to 5FIMRT, moreover, RapidArc had advantages in CI and HI of PTV, sparing of femoral head and delivery efficiency compared to 7FIMRT.

The paper is engaged in a study of the tracer of uranium leakage in the process of stope leaching.As is known,it is necessary to find the suitable tracer in the process of spray leaching so as to determine whether the leakage occurs at the bottom of the yard of uranium mine by way of in-situ blasting and stope leaching accurately and timely.For this purpose,we have chosen the most appropriate tracer and surfactant through static experiments,and,then,we have managed to simulate the tracing behavior of the tracer and the surfactant through the dynamic column experiment.The results of the aforementioned experiment have shown that the fluorescein sodium,the bromide ions,and the iodine ions prove to have no basic effect on the leaching rate of uranium whereas the distribution coefficient of the fluorescein sodium tends to be affected at the minimal degree in the mineral,which implies that the adsorption of the mineral for the fluorescein sodium turns to be minimal,too.However,it would be possible to improve the uranium leaching rate and reduce the adsorption of the mineral for the tracer with the help of the surfactant MES,DMSS,SDS,among which SDS is generally regarded to be the most effective one.For example,when the concentration of SDS is 0.008 mol/L,the uranium leaching rate can be made higher than that at other concentrations.Just as shown by the results of the dynamic experiment,the tracer fluorescein sodium itself does not basically affect the uranium ore leaching.Nevertheless,the surfactant SDS can help to increase the uranium leaching rate by 9.01％ but reduce the adsorption of the mineral for the fluorescein sodium.Particularly speaking,the initial adsorption of the mineral for the sodium can help to increase the concentration to 20.8 times as much as that of the initial penetration of fluorescein sodium in the water.What is more,we have also worked out that the distribution coefficient (Kd) of the fluorescein sodium in the uranium ore turns to be 6.66 mL/g and 4.22 mL/g under the condition with no surfactant and added surfactant SDS,the retarding factors (Rd) being 29.46,20.59,which are supposed to have been reduced by 36.63％ and 30.11％,respectively.Therefore,the leaking tracer experiment in the yard of uranium mining by the situation in-situ blasting and stope leaching we have developed can help to identify and discriminate whether there still remains leakage or not timely and accurately with the fluorescein sodium serving as tracer and the added surfactant.