The aim of this study is to evaluate the feasibility of using blood serum surface-enhanced Raman spectroscopy (SERS) to identify benign and malignant thyroid nodules. Blood serum samples collected from three different groups including healthy volunteers (n = 22), patients with benign nodules (n = 19) and malignant nodules (n = 22) were measured by SERS. The spectral analysis results demonstrate that biomolecules in serum, such as amino acids, adenine and nucleic acid bases, change differently due to the different progression of nodules. By further combining with partial least square analysis and linear discriminant analysis (PLS-LDA) method, diagnostic accuracies of 93.65% and 82.93%, sensitivities of 92.68% and 81.82% and specificities of 95.45% and 84.21% can be achieved for differentiating healthy versus thyroid nodular groups and benign versus malignant groups, respectively. The above results have suggested that the blood serum SERS technique is helpful for precise diagnosis and timely treatment for patients with thyroid nodules.

本研究的目的是评估使用血清表面增强拉曼光谱 (SERS) 识别甲状腺良恶性结节的可行性。通过SERS测量从健康志愿者（n = 22）、良性结节患者（n = 19）和恶性结节（n = 22）三个不同组收集的血清样品。光谱分析结果表明，血清中的生物分子，如氨基酸、腺嘌呤和核酸碱基，因结节进展的不同而发生不同的变化。进一步结合偏最小二乘分析和线性判别分析（PLS-LDA）方法，诊断准确率分别为93.65%和82.93%，敏感性分别为92.68%和81.82%，特异性分别为95.45%和84。区分健康与甲状腺结节组以及良性与恶性组分别可以达到 21%。以上结果提示血清SERS技术有助于甲状腺结节患者的准确诊断和及时治疗。