Highly Accurate and Robust Early Stage Detection of Cholangiocarcinoma Using Near-Lossless SERS Signal Processing with Machine Learning and 2D CNN for Point-of-care Mobile Application

ACS Omega, Journal Year: 2025, Volume and Issue: unknown

Published: March 12, 2025

Introduction: Cholangiocarcinoma (CCA), a malignancy of the bile ducts, presents significant health burden with notably high prevalence in Northeast Thailand, where its incidence ratio is 85 per 100,000 population year. The prognosis for CCA patients remains poor, particularly proximal tumors, dismal 5-year survival rate just 10%. challenge managing exacerbated by typically late detection, contributing to mortality rate. Current screening methods, such as ultrasound, are insufficient, many do not exhibit prior symptoms or detectable liver fluke (Opisthorchis viverrini: OV) infections, underscoring urgent need alternative early detection methods. Methods: In this study, we introduce novel approach utilizing surface-enhanced Raman spectroscopy (SERS) combined near-lossless signal compression via discrete wavelet transform (DWT) together 2D CNN first time. Hamster serums different stages were collected data set. DWT was employed feature extraction, enabling capture entire SERS spectrum, unlike traditional methods like PCA and LDA, which focus only on specific peaks. These features used train convolutional neural network (2D CNN), robust against translation, rotation, scaling, thus effectively addressing peak shifting issues. We validated our using gold-standard histology, notably, method could detect at an stage. ability identify stage significantly improves chances successful intervention patient outcomes. Results conclusion: Our results demonstrate that method, combining extremely compact extraction CNN, outperformed other approaches (PCA + SVM, 1D LDA achieving performance 95.1% accuracy, 95.08% sensitivity, 98.4% specificity, area under curve (AUC) 95%. trained model further deployed server mobile application interface, paving way future field experiments rural areas home-use potential point-of-care services.

Language: Английский

---

Bioorthogonal SERS-Bioluminescence Dual-Modal Imaging for Real-Time Tracking of Triple-Negative Breast Cancer Metastasis

Acta Biomaterialia, Journal Year: 2025, Volume and Issue: unknown

Published: March 1, 2025

Language: Английский

---

Artificial Intelligence-Powered Surface-Enhanced Raman Spectroscopy for Biomedical Applications

Analytical Chemistry, Journal Year: 2025, Volume and Issue: unknown

Published: March 27, 2025

Language: Английский

---

The Lung Cancer Detection and Type Determination From Plasma and Lung Tissues by Surface‐Enhanced Raman Spectroscopy

Journal of Raman Spectroscopy, Journal Year: 2025, Volume and Issue: unknown

Published: April 4, 2025

ABSTRACT The surface‐enhanced Raman spectroscopy (SERS) technique combined with chemometry can be a potential tool for early discrimination of small cell lung cancer (SCLC) and non‐small (NSCLC) from plasma and, as well the tissue samples. Based on acquired spectra applied algorithm, it is possible to distinguish between two types SCLC NSCLC associated smoking also differentiate subtypes in very fast mode comparison standard histopathology. form partial least squares regression (PLSR) discriminant analysis (PLS‐DA) method allows, first time, discriminate against tumor samples determine SCLC, NSCLC, types. presented data clearly indicate that are enough efficiently, which significantly facilitates diagnosis terms time costs analysis. Moreover, proper identification LCC samples, especially case aggressive SCLC‐like type cancer, has substantial impact patient's treatments thus may life.

Language: Английский

---

Diagnostic oriented discrimination of different Shiga toxins via PCA-assisted SERS-based plasmonic metasurface

Nanophotonics, Journal Year: 2025, Volume and Issue: unknown

Published: April 17, 2025

Abstract Plasmonic biosensors are powerful platforms for detecting various types of analytes. Specifically, surface-enhanced Raman spectroscopy (SERS) can enable label-free and selective detection. Shiga toxin-producing Escherichia coli (STEC) represents zoonotic pathogens that cause severe diseases, such as hemolytic uremic syndrome (HUS), the most important acute renal failure in children. To date, there no effective therapies STEC infection, available diagnostic methods complex inconclusive. Here, a novel nanopattern fabricated by electron beam lithography with remarkable plasmonic properties is employed SERS substrate realizing specific recognition Stx1a, Stx2a, third variation latter. A limit detection (LOD) 6.8 pM Stx1a 2 Stx2a was achieved. Our approach supported using principal component analysis (PCA) appears to be valid alternative conventional methods, allowing real-time fast situ analysis.

Language: Английский

---