A multimodal view at cancerous liver tissue by chemical bioimaging and image segmentation strategies

Six complementary biomedical imaging modalities are combined providing substantial correlation capabilities between optical, elemental and molecular tissue characteristics. Diseased rat liver tissue with hepatocellular carcinoma (HCC) was investigated, as liver cancer is a leading cause of tumor-related mortality. Bright field microscopy, autofluorescence microscopy and histological examination of hematoxylin-eosin-stained tissue displayed comprehensive optical tissue characteristics. Elemental imaging by non-destructive micro x-ray fluorescence (µXRF) and quantitative laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) revealed lower content for nine elements in HCC and lesions compared to the liver. Non-destructive quantum cascade laser (QCL)-based infrared microspectroscopy visualized higher lipid, protein, and nucleic acid content in the liver compared to HCC, and glycogen accumulations in lesions. Matrix-assisted laser desorption/ionization-trapped ion mobility spectrometry-mass spectrometry (MALDI-TIMS-MS) unveiled individual lipid distributions to characterize HCC, connective and adjacent liver tissue. Detailed image segmentation was facilitated by (bisecting) k-means clustering and interactive clustering using uniform manifold approximation and projection (UMAP) embeddings.