imATLAS · Multimodal imaging of lung tissue based on LIBS, MALDI and histology fusion to unravel sarcoidosis causes

Multimodal imaging of lung tissue based on LIBS, MALDI and histology fusion to unravel sarcoidosis causes

The imATLAS project's main objective was to develop and validate a novel multimodal imaging workflow by fusing two powerful analytical techniques: Laser-Induced Breakdown Spectroscopy (LIBS) for elemental mapping (e.g., metals) and Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MSI) for molecular mapping (e.g., lipids, peptides). The ultimate goal was to co-localize these elemental and molecular signals in human lung tissue to identify specific "fingerprints" of exposure and unravel the causes of idiopathic diseases, particularly those triggered by an exogenous agent. Identifying the specific triggering agent is critical for the diagnosis and treatment of these diseases, yet effective diagnostic tools to do this are lacking. The imATLAS project has pushed the state-of-the-art in two key areas. First, the dual-detection kHz-LIBS prototype represents a significant hardware innovation for elemental bio-imaging, solving a key limitation of standard systems. Second, the "feature-level" LIBS-MALDI fusion workflow provides a viable new method for correlating elemental and molecular data in complex tissues. The existing kHz-LIBS system was substantially re-engineered for biological analysis. A new optical microscopy path was integrated for precise focusing on transparent tissue. Most critically, a novel dual-detection (ICCD/CMOS) system was designed and implemented, allowing the system to simultaneously detect trace-level elements (like particles) and high-concentration matrix elements. The new instrument achieved a 10 µm spatial resolution—a 2x improvement on the original goal—and successfully detected metallic nanoparticles in the rodent lung models. Data processing became the project's principal focus. When a simple "low-level" fusion failed, a successful "mid-level" (or "feature-level") fusion strategy was developed. This workflow statistically correlates 2D elemental maps (from LIBS) with thousands of molecular spectra (from MALDI) to find significant co-localized signals. This work, which included the first-time application of PCA for LIBS denoising in this context, led to two peer-reviewed publications. A key result from the imATLAS project was the successful alignment and correlation of elemental LIBS images with various molecular MALDI-MSI images from a rodent tissue model, which validated this new workflow.

Data: CORDIS, © European Union