![Quantitative SERS Detection of Uric Acid via Formation of Precise Plasmonic Nanojunctions within Aggregates of Gold Nanoparticles and Cucurbit[n]uril](/_next/image?url=https%3A%2F%2Fcloudfront.jove.com%2FCDNSource%2Fteasers%2F61682.jpg&w=3840&q=50)
Machine learning-powered plasmonic pattern recognition: etch-suppressed gold nanorods for multiplex urinary analysis
Mahsa Daneshmandi1, Afsaneh Orouji1, Mohammad Reza Hormozi-Nezhad1
1Department of Chemistry, Sharif University of Technology, Tehran, 111559516, Iran. hormozi@sharif.edu.
Analytical Methods : Advancing Methods and Applications
|September 15, 2025
Summary
This study presents a novel, non-toxic multicolorimetric probe for simultaneous detection of four key catecholamine neurotransmitters (CNTs). The advanced probe utilizes gold nanorods and machine learning for accurate quantification in complex mixtures, aiding neurological disorder diagnostics.
Area of Science:
- Analytical Chemistry
- Nanotechnology
- Biomedical Engineering
Background:
- Simultaneous monitoring of catecholamine neurotransmitters (CNTs) like epinephrine (Epi), norepinephrine (NE), levodopa (L-DOPA), and dopamine (DA) is crucial for diagnosing neurological disorders.
- Existing sensing platforms often struggle with multiplexing capabilities and potential cytotoxicity, limiting their clinical applicability.
Purpose of the Study:
- To develop a sensitive, non-toxic, and non-invasive single-component multicolorimetric probe for simultaneous detection and differentiation of low concentrations of Epi, NE, L-DOPA, and DA.
- To enable quantification of binary, ternary, and quaternary mixtures of these CNTs.
- To validate the probe's performance using machine learning for robust analyte identification and quantification.
Main Methods:
- A novel sensing mechanism based on the controlled inhibition of gold nanorod (AuNR) etching by N-bromosuccinimide (NBS) was employed.
- Spectral responses were analyzed using machine learning algorithms: Linear Discriminant Analysis (LDA) for classification and Partial Least Squares Regression (PLSR) for concentration prediction.
- The probe's performance was evaluated for linearity, detection limits, and application in human urine samples.
Main Results:
- The assay demonstrated excellent linearity across wide concentration ranges for Epi, NE, L-DOPA, and DA.
- Low detection limits were achieved: 0.62 μmol L⁻¹ for Epi, 0.47 μmol L⁻¹ for NE, 0.49 μmol L⁻¹ for L-DOPA, and 0.92 μmol L⁻¹ for DA.
- The platform successfully detected and quantified CNTs in human urine samples, demonstrating practical utility.
Conclusions:
- The developed single-component multicolorimetric probe offers a sensitive, non-toxic, and non-invasive method for simultaneous CNT detection.
- The integration of machine learning enhances analyte identification and quantification accuracy.
- This platform shows significant potential for point-of-care diagnostics and clinical neurochemical analysis.
