Rapid diagnosis and severity scale of post-COVID condition using advanced spectroscopy

Abstract

The COVID-19 pandemic has resulted in a persistent health challenge known as Post-COVID Condition (PCC), characterized by symptoms lasting at least three months after the initial SARS-CoV-2 infection and potentially persisting for several years. While studies on PCC using lipidomics and proteomics have been conducted, these methods are costly and time-consuming. The comprehensive analysis of UV–VIS–NIR–MIR spectroscopy is explored here as an alternative for the rapid and cheap diagnosis and quantification of the severity of PCC. Blood samples from 65 PCC patients, previously analyzed in lipidomic and proteomic studies, along with samples from 65 new patients, were examined to develop a model that quantifies the severity of PCC based solely on spectrophotometric data. Significant spectral variability was observed in the UV–VIS region, particularly between 297 and 600 nm, correlating strongly with patient symptoms. Unsupervised clustering algorithms in this spectral region effectively differentiated between asymptomatic and symptomatic patients, achieving a Jaccard similarity score of 0.667 when compared with clinical symptom classifications. Comparative analysis with proteomic and lipidomic studies indicated that UV–VIS spectroscopy captures clinically relevant biochemical information. The results of the model developed in this work to quantify the severity of PCC demonstrated robustness with new patient data, underscoring the method’s potential as a rapid, non-invasive, and cost-effective diagnostic tool. This study highlights the strengths of spectroscopic techniques, suggesting their suitability for widespread clinical application in diagnosing and monitoring PCC, and emphasizes the need for further refinement and integration of these methods into healthcare practice, particularly for their potential implementation in portable devices.