Image Fusion and Tissue Recognition

In the last years, the research toward novel biomaterials both for orthopaedic and dental applications, has been receiving growing attention driven by the requirement to improve both the healing behaviour of tissues around implants and the quality of life in patients. Bone repair/regeneration is usually investigated through x-ray computed microtomography (mCT) supported by histology of extracted samples, to analyse biomaterial structure and new bone formation process. Magnetic Resonance Imaging (mMRI) shows a richer tissue contrast than mCT, despite at lower resolution, and could be combined with mCT in the perspective of conducting non-destructive 3D investigations of bone.

Multimodal3D developed a new pipeline combining different tomographic images to obtain a new fused 3D meta-structure of bone micromorphology.

Specifically, the multivariate clustering algorithm with the automatic selection of the cluster number allowed the data‐driven identification of different biological tissues, suggesting different types of new bone emerging during the bone maturation process. It is very important to evaluate the bone tissue heterogeneity and its distribution over the sample volume. This information allows to know how different tissues and mineral density are distributed inside the sample, which are needed to estimate the bone strength. This estimate is not possible from the measure of the average bone mineral density using only X‐ray tomogra- phy, which does not exactly disentangle between different biological tissues. On the other side, the estimation of failure risk would benefit.

Our method is especially suited for the analysis or the follow-up of biomaterial degradation and/or bone regeneration process, quantitative assessment of bone inflammatory processes and consequent bone weakness and degeneration.