Chemometric Methods Applied to Infrared and Raman Spectroscopy for Arboviruses Diagnosis: A Systematic Review With...

Area of Science:

• Biomedical Spectroscopy
• Artificial Intelligence in Diagnostics
• Infectious Disease Research

Background:

• Arboviruses (dengue, Zika, chikungunya, yellow fever) present similar symptoms, complicating diagnosis in co-circulating regions.
• Traditional lab methods for arboviruses are limited by cross-reactivity and infrastructure needs.
• Spectroscopic methods (ATR-FTIR, Raman) with AI offer rapid, cost-effective diagnostic potential.

Purpose of the Study:

• To systematically review and synthesize studies on infrared and Raman spectroscopy for arbovirus diagnosis using clinical samples and chemometric models.
• To qualitatively and quantitatively assess the diagnostic performance of spectroscopic techniques combined with AI for arboviral infections.
• To identify research gaps and the need for further validation in arbovirus diagnostics.

Main Methods:

• Systematic literature search and review (PROSPERO CRD420251006929) of studies applying spectroscopy (infrared, Raman) and AI to clinical samples.
• Analysis of 23 studies, predominantly focusing on dengue, with some including Zika and chikungunya.
• Application of multivariate analysis methods like PCA-LDA and PLS-DA.

Main Results:

• Raman spectroscopy, often paired with multivariate analysis (PCA-LDA, PLS-DA), demonstrated high diagnostic accuracy.
• Overall sensitivity for Raman spectroscopy was 0.94 (95% CI: 0.91-0.96) and specificity was 0.97 (95% CI: 0.95-0.98).
• A high risk of bias was noted across the included studies.

Conclusions:

• Spectroscopic techniques coupled with AI show significant potential for rapid and low-cost arbovirus diagnosis.
• Raman spectroscopy, in particular, exhibits promising sensitivity and specificity for detecting arboviral infections.
• Further high-quality, robust studies are essential for clinical validation and broader implementation of these advanced diagnostic approaches.