Abstract
421
Objectives Histopathological PET tracer validation frequently requires coregistration of autoradiography(AR) and images of relevant tumor microenvironment markers acquired from tissue sections.The purpose of this study is to develop an objective methodology for automated deformable coregistration of digital phosphor plate AR and microscopy images acquired from sequential tissue sections.
Methods Tumor bearing mice were injected with 18F-FLT and other markers including Hoechst (blood flow surrogate). After sacrifice, tumors were excised, frozen and sectioned. Multiple stacks of sequential 8μm sections were collected for each tumor. The middle (reference) sections were used for AR to image FLT uptake distribution. Sections adjacent to the reference were used to acquire histopathological data (pattern of cell proliferation,etc). 8μm section thickness allowed for slice-to-slice microenvironment contiguity. Hoechst images were acquired for each section. To correct for deformations and misalignments induced by tissue processing and image acquisition, Hoechst image of each non-reference section was warped onto the reference Hoechst using elastic registration.This transformation was then applied to other images acquired from the same tissue section. This way, all microscopy images were coregistered to the reference Hoechst image. The Hoechst to AR image registration was done using rigid point set registration based on external markers visible in both images.
Results Registration error was evaluated using sets of independent landmarks. The mean error of Hoechst to AR (same section) registration was 31.8μm. The error of Hoechst-based deformable registration of histopathological images was 25.2μm.Total registration error was evaluated at 40.6μm.This supersedes current rigid registration methods with reported errors of 100-200μm.
Conclusions Deformable registration of ARs and histopathology images acquired from sequential sections is feasible and highly accurate when performed using corresponding Hoechst images