Abstract

Neglected tropical diseases (NTDs) remain a major public health challenge in endemic regions due to limitations in conventional diagnostic and surveillance tools, which are often invasive, time-consuming, and resource-intensive. Here, we present an integrated photodiagnostic framework based on Fourier Transform Infrared (FTIR) spectroscopy combined with multivariate analysis and machine learning for applications spanning clinical diagnosis, pathogen identification, and vector surveillance. Blood serum FTIR spectra enabled accurate discrimination of paracoccidioidomycosis patients from healthy individuals, achieving over 90% classification accuracy using support vector machines. In experimental cutaneous leishmaniasis, FTIR analysis of biofluids and tissues revealed disease-associated molecular signatures, demonstrating the critical role of data analysis strategies in model performance. Species-level discrimination of Leishmania parasites in liquid cultures reached up to 100% accuracy, highlighting the potential of FTIR as a rapid alternative to molecular typing. Furthermore, infrared spectroscopy successfully differentiated morphologically indistinguishable female sandfly vectors with accuracies above 95%, strengthening epidemiological surveillance capabilities. Together, these studies demonstrate that FTIR spectroscopy coupled with machine learning constitutes a fast, low-cost, and scalable platform for diagnosis and monitoring of NTDs within a One Health perspective.

Biography

Cícero Cena is a Professor at the Federal University of Mato Grosso do Sul (UFMS), where he leads research in optical spectroscopy, photodiagnosis, and machine-learning-based analysis of biofluids and biological systems. His work focuses on the development of low-cost, noninvasive diagnostic and surveillance tools for infectious and neglected tropical diseases within a One-Health framework. He has authored over 60 peer-reviewed publications, with an h-index of 16.