Three-Dimensional Visualisation of Blood Vessels in Human Gliomas Using Tissue Clearing and Deep Learning

Gliomas, with their intricate and aggressive nature, call for a detailed visualisation of their vasculature. Traditional 2D imaging often overlooks the spatial heterogeneity of tumours. Our study overcomes this by combining tissue clearing, 3D-confocal microscopy imaging and deep learning-aided vessel extraction, achieving comprehensive 3D visualisation of glioma vasculature in intact human tissue. Specifically, we treated formalin-fixed thick human glioma tissue sections (500 μm) with OPTIClear for transparency and performed immunofluorescent labelling. Using confocal microscopy, we obtained 3D images of glioma vasculature. For vessel extraction, we employed a specialised 3D U-Net, enriched with image preprocessing and post-processing methods. In addition, we obtained 3D images of astrocytes or glioma cells, cell nuclei and vasculature with vascular basement membrane staining. Our findings indicated that OPTIClear-enabled tissue clearing yielded a holistic 3D representation of immunolabelled vessels and surrounding cells in human glioma samples. Our deep learning technique outperformed the traditional Imaris approach in terms of accuracy and efficiency in vessel extraction. Furthermore, discernible variations in vascular morphological metrics were observed between low- and high-grade gliomas, revealing the spatial heterogeneity of human glioma vessels. Analysis of other markers demonstrated differences in glioma cell morphology and vessel wall disruption across grades. In essence, our innovative blend of tissue clearing and deep learning not only enhances 3D visualisation of human glioma vasculature but also underscores morphological disparities across glioma grades, potentially influencing pathological grading, therapeutic strategies and prognostic evaluations.