A 2D-digital spectral sensing method for rapid antibiotic detection in water

Area of Science:

• Environmental Science
• Analytical Chemistry
• Spectroscopy

Background:

• Antibiotics are a significant global pollutant, necessitating efficient detection for ecological and health risk management.
• Traditional methods for antibiotic detection are costly and time-consuming, failing to meet rapid monitoring needs.

Purpose of the Study:

• To develop a novel, rapid, and cost-effective method for detecting antibiotics in water samples.
• To integrate spectral imaging with deep learning for enhanced antibiotic analysis.

Main Methods:

• Utilized a spectral imaging system and chemical probes to generate 2D digital spectral images (2D-DS images).
• Developed a Digital Spectral Convolutional Neural Network (DSCNN) model for end-to-end quantitative analysis of antibiotic concentrations from 2D-DS images.
• Applied the fourth research paradigm, driven by spectral big data, to analyze complex samples.

Main Results:

• The DSCNN model achieved high predictive accuracy, with R² values between 0.85 and 0.93.
• Achieved a low detection limit of 1.94 mg L⁻¹ for antibiotics.
• Developed a rapid detection platform with a single analysis time of approximately 3 minutes.

Conclusions:

• The developed method provides a high-throughput, low-cost solution for rapid antibiotic detection in water.
• This approach offers a solid technical foundation for intelligent environmental monitoring networks.
• The integration of spectral imaging and deep learning enables efficient analysis of complex environmental samples.