Multi-modal image registration and machine learning for the generation of 3D virtual histology of bone implants

Irvine, Sarah C.; Müller, Bert; Moosmann, Julian P.; Wang, Ge; Bootbool, Moral; Zeller-Plumhoff, Berit; Lucas, Christian; Galli, Silvia

doi:10.1117/12.3028465

Contribution to a conference proceedings/Journal Article

PUBDB-2024-06688

Multi-modal image registration and machine learning for the generation of 3D virtual histology of bone implants

Irvine, S. C.Extern* ; Lucas, C.Extern* ; Bootbool, M.Extern* ; Galli, S.Extern* ; Zeller-Plumhoff, B.HZG*Hereon* ; Moosmann, J. P. (Corresponding author)HZG*Hereon* ; Müller, B. (Editor)Extern* ; Wang, G. (Editor)

2024
SPIE Bellingham, Wash.

Developments in X-Ray Tomography XV, San Diego, United States, 18 Aug 2024 - 23 Aug 2024 Proceedings of SPIE 13152, 70 (2024) [10.1117/12.3028465]

This record in other databases:

Please use a persistent id in citations: doi:10.1117/12.3028465

Abstract: In our correlative characterisation studies of biodegradable and permanent metal bone implants, we have performed both synchrotron-radiation microtomography (SR-μCT) and histology on the same samples. Histological staining is still the gold standard for tissue visualisation yet requires multiple time-consuming sample preparation steps (fixing, embedding, sectioning and staining) before imaging is performed on individual slices, in contrast to the non-invasive and 3D nature of tomography. In the process of correlating the corresponding data sets, we are able to combine advantages of both modalities by using machine learning methods to generate artificially stained 3D virtual histology datasets from SR-μCT datasets. For this we have developed an automated registration tool to find and fit the correct virtual tomographic plane to each histology slice. Preliminary results are promising after training a modified cycle generative adversarial network on our data, with two different histological stainings.

Classification: