Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Jul 31;14(8):372.
doi: 10.3390/bios14080372.

Label-Free Surface-Enhanced Raman Spectroscopy with Machine Learning for the Diagnosis of Thyroid Cancer by Using Fine-Needle Aspiration Liquid Samples

Lili Gao  1 , Siyi Wu  2 , Puwasit Wongwasuratthakul  2 , Zhou Chen  2 , Wei Cai  3 , Qinyu Li  3 , Linley Li Lin  2
Affiliations

Label-Free Surface-Enhanced Raman Spectroscopy with Machine Learning for the Diagnosis of Thyroid Cancer by Using Fine-Needle Aspiration Liquid Samples

Lili Gao et al. Biosensors (Basel). .

Abstract

The incidence of thyroid cancer is increasing worldwide. Fine-needle aspiration (FNA) cytology is widely applied with the use of extracted biological cell samples, but current FNA cytology is labor-intensive, time-consuming, and can lead to the risk of false-negative results. Surface-enhanced Raman spectroscopy (SERS) combined with machine learning algorithms holds promise for cancer diagnosis. In this study, we develop a label-free SERS liquid biopsy method with machine learning for the rapid and accurate diagnosis of thyroid cancer by using thyroid FNA washout fluids. These liquid supernatants are mixed with silver nanoparticle colloids, and dispersed in quartz capillary for SERS measurements to discriminate between healthy and malignant samples. We collect Raman spectra of 36 thyroid FNA samples (18 malignant and 18 benign) and compare four classification models: Principal Component Analysis-Linear Discriminant Analysis (PCA-LDA), Random Forest (RF), Support Vector Machine (SVM), and Convolutional Neural Network (CNN). The results show that the CNN algorithm is the most precise, with a high accuracy of 88.1%, sensitivity of 87.8%, and the area under the receiver operating characteristic curve of 0.953. Our approach is simple, convenient, and cost-effective. This study indicates that label-free SERS liquid biopsy assisted by deep learning models holds great promise for the early detection and screening of thyroid cancer.

Keywords: CNN; Raman spectroscopy; liquid biopsy; machine learning; thyroid fluid.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflicts of interest.

Figures

Figure 1
Figure 1
Scheme of the work flow. FNA samples are first used for cell cytology, as the gold standard diagnosis. Then, excess FNA samples are stored in the preservation phase as the FNA washout fluids. The fluid supernantant is mixed with Ag colloids for SERS enhancement, and this liquid mixture is loaded into microfluidic quartz capillaries for Raman measurements. We collected the Raman spectra for machine learning classification. Cell cytology results are used for evaluation of SERS diagnosis accuracy.
Figure 2
Figure 2
Workflow of the sample preparation and SERS measurements.
Figure 3
Figure 3
(a) Group division. (b) CNN architecture and hyperparameters.
Figure 4
Figure 4
Label-free SERS measurements using the Ag colloids and thyroid FNA washout liquids. (a) TEM image and (b) UV-Vis spectra of the Ag colloids. (c) Picture of the SERS measurement setup. (d) SERS measurements of the pure preservation buffer mixed with Ag colloids. The mixture was measured without or with open-air exposure for 5 or 10 min with mild heating. The spectrum of pure ethanol (black curve) is for comparison. (e) SERS measurements of the FNA washout fluid supernatant mixed with Ag colloids. Representative spectra of a benign or a malignant sample were displayed. The mixture was measured without exposure or after 5 min open-air exposure with mild heating.
Figure 5
Figure 5
FNA cytology results and label-free SERS measurements of FNA washout liquids. (a) FNA cell cytology image. (b) Averaged SERS spectra with standard deviations of thyroid malignant and benign samples, as well as the averaged difference spectrum (malignant–benign). This is the label-free SERS measurements of the FNA washout liquids mixed with Ag colloids. Standard deviation values are shown in shadow. The SERS spectrum of the preservation buffer is also displayed (black curve).
Figure 6
Figure 6
(a) PCA clustering results. (b) Confusion matrix displaying the classification performance using PCA-LDA, RF, SVM, and CNN.
Figure 7
Figure 7
(a) ROC curves of dichotomous discriminations between the thyroid malignant and benign samples using the four models. The corresponding AUC for the CNN model is 0.953. The dotted line indicates performance under random events (0.5 AUC). (b) The characteristic Raman peak contribution in the CNN model.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Paschou S.A., Vryonidou A., Goulis D.G. Thyroid nodules: A guide to assessment, treatment and follow-up. Maturitas. 2017;96:1–9. doi: 10.1016/j.maturitas.2016.11.002. - DOI - PubMed
    1. Luo J., Zhang C., Huang F., Chen J., Sun Y., Xu K., Huang P. Risk of malignancy in thyroid nodules: Predictive value of puncture feeling of grittiness in the process of fine-needle aspiration. Sci. Rep. 2017;7:13109. doi: 10.1038/s41598-017-13391-3. - DOI - PMC - PubMed
    1. Cibas E.S., Ali S.Z. The 2017 Bethesda System for Reporting Thyroid Cytopathology. Thyroid. 2017;27:1341–1346. doi: 10.1089/thy.2017.0500. - DOI - PubMed
    1. Landa I., Cabanillas M.E. Genomic alterations in thyroid cancer: Biological and clinical insights. Nat. Rev. Endocrinol. 2024;20:93–110. doi: 10.1038/s41574-023-00920-6. - DOI - PubMed
    1. Huang L., Sun H., Sun L., Shi K., Chen Y., Ren X., Ge Y., Jiang D., Liu X., Knoll W., et al. Rapid, label-free histopathological diagnosis of liver cancer based on Raman spectroscopy and deep learning. Nat. Commun. 2023;14:48. doi: 10.1038/s41467-022-35696-2. - DOI - PMC - PubMed

LinkOut - more resources