Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells

E. Ping Tsai; I. Hsuan Chou; Yen Pei Lin; Sanskruti Swain; Ting Yi Lin; Yun Jung Ting; Hsing Ying Lin; Chen Han Huang

doi:10.1117/12.3044069

Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells

E. Ping Tsai
, I. Hsuan Chou
, Yen Pei Lin
, Sanskruti Swain
, Ting Yi Lin
, Yun Jung Ting
, Hsing Ying Lin
, Chen Han Huang

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Surface-enhanced Raman scattering (SERS) tags, combining nanoparticles with Raman reporter molecules, offer high sensitivity and quantification for detecting target molecules. However, analyzing SERS spectra, especially in mixtures, is challenging due to interference and noise. This study developed a Convolutional Neural Network-based model trained on 10,000 augmented SERS spectra to identify components in mixtures. Baseline correction using airPLS and the Nonnegative Elastic Net algorithm enabled accurate ratio estimation of each reporter. The system used antibody-functionalized SERS tags for multiplex labeling and AI spectral analysis for component classification and concentration calculation in single cancer cells, showing high detection specificity. This method significantly enhances multispectral recognition, promising advances in tumor cell analysis.

Original language	English
Title of host publication	Plasmonics in Biology and Medicine XXII
Editors	Tuan Vo-Dinh, Ho-Pui A. Ho, Krishanu Ray
Publisher	SPIE
ISBN (Electronic)	9781510684225
DOIs	https://doi.org/10.1117/12.3044069
State	Published - 2025
Event	Plasmonics in Biology and Medicine XXII 2025 - San Francisco, United States Duration: 26 Jan 2025 → 28 Jan 2025

Publication series

Name	Progress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume	13337
ISSN (Print)	1605-7422

Conference

Conference	Plasmonics in Biology and Medicine XXII 2025
Country/Territory	United States
City	San Francisco
Period	26/01/25 → 28/01/25

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Convolutional Neural Network
Raman
SERS tags

Access to Document

10.1117/12.3044069

Cite this

Tsai, E. P., Chou, I. H., Lin, Y. P., Swain, S., Lin, T. Y., Ting, Y. J., Lin, H. Y., & Huang, C. H. (2025). Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells. In T. Vo-Dinh, H.-P. A. Ho, & K. Ray (Eds.), Plasmonics in Biology and Medicine XXII Article 1333702 (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 13337). SPIE. https://doi.org/10.1117/12.3044069

@inproceedings{15369a954efd4b5fb421e819118fdb6e,

title = "Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells",

abstract = "Surface-enhanced Raman scattering (SERS) tags, combining nanoparticles with Raman reporter molecules, offer high sensitivity and quantification for detecting target molecules. However, analyzing SERS spectra, especially in mixtures, is challenging due to interference and noise. This study developed a Convolutional Neural Network-based model trained on 10,000 augmented SERS spectra to identify components in mixtures. Baseline correction using airPLS and the Nonnegative Elastic Net algorithm enabled accurate ratio estimation of each reporter. The system used antibody-functionalized SERS tags for multiplex labeling and AI spectral analysis for component classification and concentration calculation in single cancer cells, showing high detection specificity. This method significantly enhances multispectral recognition, promising advances in tumor cell analysis.",

keywords = "Convolutional Neural Network, Raman, SERS tags",

author = "Tsai, \{E. Ping\} and Chou, \{I. Hsuan\} and Lin, \{Yen Pei\} and Sanskruti Swain and Lin, \{Ting Yi\} and Ting, \{Yun Jung\} and Lin, \{Hsing Ying\} and Huang, \{Chen Han\}",

note = "Publisher Copyright: {\textcopyright} 2025 SPIE.; Plasmonics in Biology and Medicine XXII 2025 ; Conference date: 26-01-2025 Through 28-01-2025",

year = "2025",

doi = "10.1117/12.3044069",

language = "???core.languages.en\_GB???",

series = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",

publisher = "SPIE",

editor = "Tuan Vo-Dinh and Ho, \{Ho-Pui A.\} and Krishanu Ray",

booktitle = "Plasmonics in Biology and Medicine XXII",

}

Tsai, EP, Chou, IH, Lin, YP, Swain, S, Lin, TY, Ting, YJ, Lin, HY & Huang, CH 2025, Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells. in T Vo-Dinh, H-PA Ho & K Ray (eds), Plasmonics in Biology and Medicine XXII., 1333702, Progress in Biomedical Optics and Imaging - Proceedings of SPIE, vol. 13337, SPIE, Plasmonics in Biology and Medicine XXII 2025, San Francisco, United States, 26/01/25. https://doi.org/10.1117/12.3044069

Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells. / Tsai, E. Ping; Chou, I. Hsuan; Lin, Yen Pei et al.
Plasmonics in Biology and Medicine XXII. ed. / Tuan Vo-Dinh; Ho-Pui A. Ho; Krishanu Ray. SPIE, 2025. 1333702 (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 13337).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells

AU - Tsai, E. Ping

AU - Chou, I. Hsuan

AU - Lin, Yen Pei

AU - Swain, Sanskruti

AU - Lin, Ting Yi

AU - Ting, Yun Jung

AU - Lin, Hsing Ying

AU - Huang, Chen Han

PY - 2025

Y1 - 2025

N2 - Surface-enhanced Raman scattering (SERS) tags, combining nanoparticles with Raman reporter molecules, offer high sensitivity and quantification for detecting target molecules. However, analyzing SERS spectra, especially in mixtures, is challenging due to interference and noise. This study developed a Convolutional Neural Network-based model trained on 10,000 augmented SERS spectra to identify components in mixtures. Baseline correction using airPLS and the Nonnegative Elastic Net algorithm enabled accurate ratio estimation of each reporter. The system used antibody-functionalized SERS tags for multiplex labeling and AI spectral analysis for component classification and concentration calculation in single cancer cells, showing high detection specificity. This method significantly enhances multispectral recognition, promising advances in tumor cell analysis.

AB - Surface-enhanced Raman scattering (SERS) tags, combining nanoparticles with Raman reporter molecules, offer high sensitivity and quantification for detecting target molecules. However, analyzing SERS spectra, especially in mixtures, is challenging due to interference and noise. This study developed a Convolutional Neural Network-based model trained on 10,000 augmented SERS spectra to identify components in mixtures. Baseline correction using airPLS and the Nonnegative Elastic Net algorithm enabled accurate ratio estimation of each reporter. The system used antibody-functionalized SERS tags for multiplex labeling and AI spectral analysis for component classification and concentration calculation in single cancer cells, showing high detection specificity. This method significantly enhances multispectral recognition, promising advances in tumor cell analysis.

KW - Convolutional Neural Network

KW - Raman

KW - SERS tags

UR - https://www.scopus.com/pages/publications/105004317934

U2 - 10.1117/12.3044069

DO - 10.1117/12.3044069

M3 - 會議論文篇章

AN - SCOPUS:105004317934

T3 - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

BT - Plasmonics in Biology and Medicine XXII

A2 - Vo-Dinh, Tuan

A2 - Ho, Ho-Pui A.

A2 - Ray, Krishanu

PB - SPIE

T2 - Plasmonics in Biology and Medicine XXII 2025

Y2 - 26 January 2025 through 28 January 2025

ER -

Deep-Learning-Enhanced SERS Tag Analysis for Multiplex Detection of Cancer Cells

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Fingerprint

Cite this