Abstract
Examination of histological or immunohistochemically stained 2D sections of embedded tissue is one of the most frequently used tools in biomedical research and clinical routine. Since to date, targeted sectioning of specific regions of interest (ROI) in the sample is not possible, we aimed at developing a guided sectioning approach based on x-ray 3D virtual histology for heavy ion stained murine lung samples. For this purpose, we increased the contrast to noise ratio of a standard benchtop microCT by 5-10-fold using free-propagation phase contrast imaging and thus substantially improved image quality. We then show that microCT 3D datasets deliver more precise anatomical information and quantification of the sample than traditional histological sections, which display deformations of the tissue. To quantify these deformations caused by sectioning we developed the "Displacement Index (DI)", which combines block-matching with the calculation of the local mutual information. We show that the DI substantially decreases when a femtosecond laser microtome is used for sections as opposed to a traditional microtome. In conclusion, our microCT based virtual histology approach can be used as a supplement and a guidance tool for traditional histology, providing 3D measurement capabilities and offering the ability to perform sectioning directly at an ROI.