Development of a SERS based cancer diagnosis approach employing cryosectioned thyroid tissue samples on PDMS

An efficient SERS based novel analytical approach named Cryosectioned-PDMS was developed systematically and evaluated applying on 64 thyroid biopsy samples. To utilize thyroid biopsy samples, a 20-mu l volume of h-AgNPs suspension was dropped on a 5-mu m thick cryosectioned biopsy specimen placed on the PDMS coated glass slide. The SERS spectra from a 10 x 10 points array acquired by mapping 22.5 mu m x 22.5 mu m sized area from suspended dried droplets placed on the tissue surface. The probability of correctly predicted performance for diagnosis of malignant, benign and healthy tissues was resulted in the accuracy of 100 % for the spectral bands at 667, 724, 920, 960, 1052, 1096, 1315 and 1457 cm(-1) using PCA-fed LDA machine learning. The Cryosectioned-PDMS biophotonic approach with PCA-LDA predictive model demonstrated that the vibrational signatures can accurately recognize the fingerprint of cancer pathology from a healthy one with a simple and fast sample preparation methodology. (C) 2022 Elsevier Inc. All rights reserved.

Automated summary

An efficient SERS-based analytical approach named Cryosectioned-PDMS was developed and evaluated using thyroid biopsy samples. By mapping the SERS spectra from suspended droplets on the tissue surface, the vibrational signatures were successfully utilized to accurately diagnose malignant, benign, and healthy tissues with a 100% accuracy rate using machine learning.