Recent advances in optical spectroscopy have unveiled new opportunities for probing living systems at a molecular level. Our primary objective is to advance and assess vibrational spectroscopy as an analytical framework for comprehensive cross-molecular profiling of systemic human biofl uids and to evaluate the feasibility of employing infrared fingerprinting for high-throughput in vitro biomedical diagnostics. More specifically, we are investigating its viability for the analysis of human blood serum and plasma within the context of clinical diagnostics.

By combining vibrational fingerprinting of liquid blood plasma and serum with the integration of data analyses using machine learning, the results based of various studies^1,2,3,4 will be discussed. On a smaller scale, we have uncovered a remarkable degree of stability in the infrared molecular fi ngerprints within individuals over time¹, laying a crucial foundation for potential health screening applications.

In large scale case-control clinical study, we have gathered evidence indicating that the spectral information of plasma and serum contains distinct signatures for four types of cancer². The effectiveness of disease detection is linked to the stage of cancer progression. We underscore the potential for early cancer diagnostics, highlight possible applications forging primary cancer diagnostics and provide evidence for distinguishing between different cancer entities. Yet, blood-based vibrational fingerprints are not able to detect only very severe health phenotypes like cancer. Infrared spectra also carry the capacity to just as well detect common chronic health deviations⁵. Interestingly, in a large-scale population scenario with its inherent heterogeneity, we identified further potential of infrared fingerprinting to distinguish between various co-occurring conditions, opening up the possibility of screening for a variety of conditions and enhancing risk stratification.