Rapid diagnosis of bacterial vaginosis using machine-learning-assisted surface-enhanced Raman spectroscopy of human vaginal fluids

Abstract

Bacterial vaginosis (BV) is an abnormal gynecological condition caused by the overgrowth of specific bacteria in the vagina. This study aims to develop a novel method for BV detection by integrating surface-enhanced Raman scattering (SERS) with machine learning (ML) algorithms. Vaginal fluid samples were classified as BV positive or BV negative using the BVBlue Test and clinical microscopy, followed by SERS spectral acquisition to construct the data set. Preliminary SERS spectral analysis revealed notable disparities in characteristic peak features. Multiple ML models were constructed and optimized, with the convolutional neural network (CNN) model achieving the highest prediction accuracy at 99%. Gradient-weighted class activation mapping (Grad-CAM) was used to highlight important regions in the images for prediction. Moreover, the CNN model was blindly tested on SERS spectra of vaginal fluid samples collected from 40 participants with unknown BV infection status, achieving a prediction accuracy of 90.75% compared with the results of the BVBlue Test combined with clinical microscopy. This novel technique is simple, cheap, and rapid in accurately diagnosing bacterial vaginosis, potentially complementing current diagnostic methods in clinical laboratories.

Importance: The accurate and rapid diagnosis of bacterial vaginosis (BV) is crucial due to its high prevalence and association with serious health complications, including increased risk of sexually transmitted infections and adverse pregnancy outcomes. Although widely used, traditional diagnostic methods have significant limitations in subjectivity, complexity, and cost. The development of a novel diagnostic approach that integrates SERS with ML offers a promising solution. The CNN model's high prediction accuracy, cost-effectiveness, and extraordinary rapidity underscore its significant potential to enhance the diagnosis of BV in clinical settings. This method not only addresses the limitations of current diagnostic tools but also provides a more accessible and reliable option for healthcare providers, ultimately enhancing patient care and health outcomes.

Keywords: SERS; bacterial vaginosis; deep learning; machine learning; rapid identification.

Fig 1

Structure and properties of silver nanoparticles. (A) TEM image of AgNPs. (B) The absorption spectrum of AgNPs; the characteristic absorption peak is located at 428 nm. (C) Histogram of the diameter distribution of silver nanoparticles. The red bars in the figure represent the number distribution of particles with different diameters, and the covered black curve is the Gaussian fit of the distribution, with an SD of 26.37 nm, used to describe the width of the particle size distribution. (D) SERS spectral curves of different solutions. Each curve represents the Raman scattering intensity as a function of Raman shift for a specific compound.

Fig 2

SERS and PCA analysis. (A and B) SERS fingerprints of BV-positive and BV-negative samples. (C and D) Deconvolution spectra of positive and negative SERS signals. (E and F) Loading plots for positive and negative SERS signals.

Fig 3

Comparison of OPLS-DA performance between the raw and normalized SERS data. (A) Raw and (B) normalized BV-positive and BV-negative scatter plots. R2X and R2Y represent the percentage of input matrix information that the model can explain respectively. Q2 is used to evaluate the predictive ability of the model. The range of three metrics is (0, 1), and the larger the value, the better the prediction performance of the model.

Fig 4

Performance evaluation of ML models. (A) ROC curves of seven algorithms and ROC curves of different colors have been quantified with AUC values. (B) Confusion matrix of CNN. TPR, true positive rate; FPR, false positive rate.

Fig 5

CNN architecture and classification activation mapping. (A) Heatmaps generated by Grad-CAM technology. (B) Structure diagram of CNN.

Fig 6

Blind sample testing of CNN models. (A) BV-positive prediction distribution. (B) BV-negative prediction distribution.