Deep Learning Assisted Surface-Enhanced Raman Spectroscopy (SERS) for Rapid and Direct Nucleic Acid Amplification and Detection: Toward Enhanced Molecular Diagnostics.

Researchers

Journal

Modalities

Models

Abstract

Surface-enhanced Raman scattering (SERS) has evolved into a robust analytical technique capable of detecting a variety of biomolecules despite challenges in securing a reliable Raman signal. Conventional SERS-based nucleic acid detection relies on hybridization assays, but reproducibility and signal strength issues have hindered research on directly amplifying nucleic acids on SERS surfaces. This study introduces a deep learning assisted ZnO-Au-SERS-based direct amplification (ZADA) system for rapid, sensitive molecular diagnostics. The system employs a SERS substrate fabricated by depositing gold on uniformly grown ZnO nanorods. These nanorods create hot spots for the amplification of the target nucleic acids directly on the SERS surface, eliminating the need for postamplification hybridization and Raman reporters. The limit of detection of the ZADA system was superior to those of the conventional amplification methods. Clinical validation of the ZADA system with coronavirus disease 2019 (COVID-19) samples from human patients yielded a sensitivity and specificity of 92.31% and 81.25%, respectively. The integration of a deep learning program further enhanced sensitivity and specificity to 100% and reduced SERS analysis time, showcasing the potential of the ZADA system for rapid, label-free disease diagnosis via direct nucleic acid amplification and detection within 20 min.