Detection of Narcotic Drugs in Urine Samples Using Attenuated Total Reflectance-Fourier Transform Infrared...

Area of Science:

• Analytical Chemistry
• Forensic Science
• Computational Biology

Background:

• Drug abuse presents a significant global health challenge, with current urine drug detection methods being slow and requiring expert analysis.
• Rapid and accurate screening for illicit substances in biological samples is crucial for effective intervention and public health.

Purpose of the Study:

• To develop and validate a rapid screening method for methamphetamines and tetrahydrocannabinol in urine using ATR-FTIR spectroscopy and machine learning.
• To differentiate between drug users and non-drug users based on urine spectral data.

Main Methods:

• Utilized attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy to analyze urine samples.
• Employed machine learning algorithms, specifically support vector machine (SVM), for classification.
• Optimized sample preparation by using urine supernatant for clearer spectral data.

Main Results:

• Urine supernatant provided better spectral discrimination than whole urine samples.
• Principal component analysis identified key spectral regions (3500-3000 cm^-1) for model development.
• A support vector machine model achieved 80.8% accuracy, 80.8% sensitivity, and 85.2% specificity in classifying drug users' urine samples.

Conclusions:

• ATR-FTIR spectroscopy combined with machine learning offers a promising approach for rapid drug screening in urine.
• The developed SVM model demonstrates suitability for classifying drug and non-drug users.
• Further research with larger sample sizes and comprehensive spectral libraries is recommended to enhance model accuracy and applicability.