肺癌是世界范围内最常见的恶性肿瘤之一，目前关于肺癌早期诊断的有效筛选方法仍然很匮乏。异戊二烯是肺癌的潜在呼吸生物标记物。无创的呼吸分析为肺癌早期诊断提供了有益的思路。光腔衰荡光谱（cavity ringdown spectroscopy，CRDS）是一种可用于定量分析的探测灵敏度极高的激光光谱技术，相比于质谱技术，它具有便携、成本低的特性，相比于电化学传感器，具有高灵敏、高精度的特点，对于呼气分子的快速、准确、高效的测量具有独特优势，这些优点使得CRDS十分适合于人体呼气异戊二烯的实时在线测量。

目的：建立能满足临床检测极限（limit of detection，LoD)要求的呼气异戊二烯CRDS测量方法，搭建基于CRDS呼气异戊二烯分析系统，实现对人体呼气异戊二烯的高准确性、高稳定性、高灵敏度分析，并初步探索CRDS的双分子协同测量方法的可行性。推进CRDS技术在无创肺癌诊断和筛查中的应用。

方法：首先筛选出激光波长—226.56 nm，并进行了异戊二烯吸收光谱干扰分析。使用调Q型Nd:YAG泵浦激光器泵浦的染料激光器作为激光光源产生紫外激光。搭建了呼气异戊二烯分析系统，对典型的CRDS衰荡信号进行了线性拟合。使用异戊二烯标准气体测量了在223-227.4 nm的吸收截面。对系统性能进行了评估，如稳定性、LoD、线性响应、重复性、响应速度、准确性。研究了气体采样袋、水蒸气、背景空气对实验结果的影响。随后，对比了74例肺癌患者、15例肺部良性病变患者和58例健康志愿者的呼气异戊二烯水平，并评价呼气异戊二烯对于肺癌诊断的意义。最后，测试了基于CRDS的双分子协同测量方法的可行性。

结果：1.光谱干扰分析表明在226.56 nm存在最小甚至近似为零的光谱干扰，因此选择226.56 nm作为最佳激光波长。2.CRDS呼气异戊二烯分析系统典型的真空衰荡信号显示出良好的线性（R²=0.99）。真空衰荡时间常数为0.88624 ，稳定性为0.2%，镜片反射率为99.81％，LoD为0.47 ppbv。系统具有线性响应（R²=0.99）和高准确性（R²=0.906）。3.异戊二烯在223-227.4 nm范围内吸收截面随着激光波长的增加而逐渐减小，与文献报道一致。异戊二烯在226.56 nm处的吸收截面为(1.93±0.0173) × 10^-17cm²/molecule。4.FEP气体采样袋具有更好的时间稳定性。水蒸气对测量结果无影响。选取了实验室外空气作为背景空气。5.健康人的呼气异戊二烯水平（274.1±130.4 ppbv）显著高于肺癌患者（114.9±38.4 ppbv）和肺部良性病变患者（128.9±43.2 ppbv）的呼气异戊二烯浓度。三组受试者的呼气异戊二烯水平均不受性别、年龄、BMI的影响。ROC曲线结果证明呼气异戊二烯浓度对肺癌的诊断有意义（AUC=0.915，敏感度为82.8%，特异性为89.2%，P<0.01）。6.基于CRDS的双分子协同测量方法测量得到的一氧化氮（nitric oxide, NO）的相对误差的绝对值在0.2-7.2%，异戊二烯的相对误差的绝对值在0-9.2%。

结论：呼气异戊二烯CRDS测量方法是对人体呼气异戊二烯进行高准确性、高稳定性、高灵敏度的可靠方法，可以实现对肺癌患者以及健康人呼气的无创检测，进一步为肺癌的早期诊断和疾病筛查提供依据。

Lung cancer is one of the most common malignant tumors worldwide. Currently, effective screening methods for early lung cancer are still scarce. Isoprene is a potential breath biomarker for lung cancer. Non-invasive breath analysis provides useful idea for early diagnosis of lung cancer. Cavity ringdown spectroscopy (CRDS) is a laser spectroscopy technology with high sensitivity for quantitative analysis. Compared with mass spectrometry, CRDS is portable and low cost. Compared with electrochemical sensors, CRDS has high sensitivity and high precision. CRDS has unique advantages for real time, accurate, and efficient measurement of breath biomarkers. These advantages make CRDS very suitable for real time and online measurement of human breath isoprene. 

Objective: To establish a CRDS measurement method for breath isoprene that can meet the requirements of clinical limit of detection (LoD). To build a CRDS-based breath isoprene analysis system and achieve high accuracy, high stability, high sensitivity analysis of human breath isoprene. To explore the feasibility of the two-molecule cooperative measurement method of CRDS preliminarily. And to promote the application of CRDS in non-invasive lung cancer diagnosis and and screening.

Methods: First, the laser wavelength-226.56 nm is selected, and the interference analysis of the isoprene absorption spectrum is discussed.A dye laser pumped bya Q-switch Nd:YAG is utilized as the light source to generate the UV light. A CRDS-based breath isoprene analysis systemis built, and the typical CRDS ringdown signal is linearly fitted.The absorption cross section at 223-227.4 nm is measured using isoprene standard gas.Evaluate system performance, including stability, detection limit, linear response, repeatability,time response, and accuracy. The influence of breath sample bags, water vapor, and background air on the experimental results are studied. Subsequently, the concentrations of breath isoprene in 74 patients with lung cancer, 15 patients with benign lung lesions and 58 healthy volunteers are compared. The significance of breath isoprene in the diagnosis of lung cancer is evaluated. Finally, the feasibility of the CRDS-based two-molecule cooperative measurement method is tested.

Results: 1. We investigate spectral absorption interference and selected an optimal detection wavelength at 226.56 nm with minimum to no spectral interference. 2. The fitting result of typical ringdown waveform of CRDS-based breath isoprene analysis system recorded under vacuum shows good linearity (R²=0.99). The ringdown time under vacuum is 0.88624  with stability of 0.2％, which leads to mirror reflectivity of 99.81% and LoD of 0.47 ppbv. The CRDS system shows good linear response (R²=0.99), and high accuracy (R²=0.906). 3.The decreasing trend of absorption cross-section of isoprene at 223-227.4 nmis also in a good agreement with those reported. The absorption cross-section at 226.56 nm is (1.93±0.0173) × 10^-17cm²/molecule.4. FEP bag has better stability over time. Water vapor has no effect on results. The air outside the laboratory is selected as the background air. 5.Breath isoprene concentration is significantly higher in healthy individuals (274.1±130.4 ppbv) than in patients with lung cancer (114.9±38.4 ppbv) and benign lung lesions (128.9±43.2 ppbv). Breath isoprene concentrations of the three groups of subjects are not affected by gender, age, and BMI. ROC curve result proves that the concentration of breath isoprene is meaningful for the diagnosis of lung cancer (AUC=0.915, sensitivity 82.8%, specificity 89.2%, P<0.01). 6. The absolute value of the relative error of NO measured by the CRDS-based two-molecule cooperative measurement method is 0.2-7.2%, and the absolute value of the relative error of isoprene is 0-9.2%.

Conclusion: The CRDS measurement method of breathisoprene is a reliable method with high accuracy, high stability and high sensitivity for human breathisoprene, which can achieve non-invasive detection of breath from lung cancer patients and healthy people, and further provide a basis for early diagnosis and disease screening of lung cancer.