Label-free analysis of experimental lung fibrosis using multiphoton microscopy and Raman spectroscopy

Abstract

<b>Introduction:</b> Chronic lung diseases such as idiopathic pulmonary fibrosis (IPF) represent a major cause of morbidity and mortality. Despite a reduced need for tissue biopsy in diagnosing IPF, conventional histology based on artificial tissue staining and scoring still represents the indispensable gold standard. <b>Rationale:</b> The emerging field of label-free biophotonics, however, has made indispensable contributions toward refining contrast mechanisms by exploiting intrinsic light-tissue interactions to generate scientific and diagnostic readout without a need for tissue preparation and/or exogenous markers. <b>Methods:</b> While label-free Raman spectroscopy (RS) is able to analyse the biochemical fingerprint of a sample, label-free multiphoton microscopy (MPM) unveils 3D tissue morphology, native autofluorescence signature, and optical polarization properties. However, both modalities within a single system are not widely available. We designed a multimodal experimental platform integrating RS and MPM with additional data analysis using machine learning. We carefully evaluated our novel approach by comparison to conventional quantification methods in inflammatory fibrogenesis of the murine lung using topical bleomycin. Results/Conclusion: Our current data show that the technical realisation of our approach is feasible and validly quantifies lung fibrosis and inflammatory cell infiltration. We hypothesize that the multi-modal data sets generated by the platform, can provide new scientific insights into inflammatory processes and tissue remodelling in lung diseases. Moreover, such data might enable more accurate diagnosis by cross-correlations of data across modalities in the future.