MO‐FF‐A3‐06: Preliminary Feasibility Study: Modeling 3D Deformations of the Prostate From Whole‐Mount Histology to in Vivo MRI

Abstract

Purpose: To investigate the accuracy of a 3D biomechanical model‐based deformation algorithm (MORFEUS) in modeling the prostate deformation that occurs between in vivo magnetic resonance imaging (MRI) and identification of the tumor on whole‐mount histology. Method and Materials: Three image sets were acquired for 10 patients: 1) in vivo T2‐weighted MR images acquired prior to prostatectomy, 2) ex vivo T2‐weighted MR images, and 3) digital images of the histological slices, rigidly registered to construct a 3D volumetric image. All three images sets were imported into the radiation treatment planning system for contouring. The entire prostate gland, the peripheral zone and central gland were contoured. The prostate was converted into a finite element model, where each zone was assigned the appropriate material property. Naturally occurring structural and morphological features (e.g. urethra) were identified as verification points in the in vivo, ex vivo, and histological images, for quantification of the accuracy of the deformable registration. MORFEUS was used to model the deformations that occur due to excision and fixation either directly, deforming histology to in vivo MRI, or using a two‐step process, histology to in vivo MRI via an intermediate step, using ex vivo MRI. Results: Initial analysis has been completed for a subset of the patients. Uncertainties following rigid registration alone exceeded 8.0mm. No significant improvements were observed when including the intermediate deformation step. The average absolute error following deformable registration, based on the verification points, was 1.3, 1.2, and 1.9mm in the left/right, anterior/posterior, and superior/inferior directions, respectively. This error is smaller than the 3 mm image slice thickness. Conclusions: Substantial deformation confounds the ability to compare histology with in vivo imaging. Deformable registration using MORFEUS can be used to resolve the deformation, enabling quantitative evaluation of in vivo imaging based on histology as a gold standard for tumor definition.

Quasi‐static magnetic resonance elastography at 7 T to measure the effect of pathology before and after fixation on tissue biomechanical properties

“… FEM iterative algorithm (showing data from a simulation experiment): (1) finite elements and boundary conditions defined, and MRE used to measure strain; (2) strain used to estimate the initial E values; (3) ABAQUS software simulates the compression and calculates three components of stress per element; (4) MRE‐measured strain combined with calculated stress to provide new estimates of E and the gel E values are reset to their average, and convergence tested; (5) once converged, values are divided by the average gel value to obtain ratio‐ E values; and (6) multiplication by the indentation‐measured gel E value to obtain quantitative E measures for tissue elements. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]…”

“…To determine sensitivity and specificity, parameters must be evaluated against the gold standard of histology (3). Whole‐mount histopathology of radical prostatectomy specimens maps the spatial distribution of disease across the tissue cross section (3) and can be reconstructed into a 3D digital histopathology volume with the tumor as a subvolume (4). As prostate tissue undergoes deformation and shrinkage between the in vivo and histopathology states, a deformable registration approach is required for accurate correlation (5).…”

“…This involves splitting the organ volume into a 3D mesh of finite elements with individually defined material properties, which is morphed from its initial shape to the target shape through a boundary surface projection. Biomechanical‐based registration is easily extended to the histopathology correlation task (4, 7, 8). However, for a step‐wise implementation through the ex vivo tissue volumes, 3D high‐resolution information is required on the ex vivo biomechanical properties, and the changes that occur in these properties as a result of fixation (9).…”

Biomechanical model-based deformable registration of MRI and histopathology for clinical prostatectomy

“…McNiven et al . [ 12 ] used a biomechanical-based registration to match histology images of the prostate with in vivo MRI. Appropriate material property models derived from the literature were used in this initial investigation.…”

Technical Note: Method to correlate whole‐specimen histopathology of radical prostatectomy with diagnostic MR imaging

“…The data obtained will be used in 3-D finite element modeling (FEM) biomechanical-based deformable registration [12][13][14] of the histopathology volume to the in vivo volume in a stepwise process through the tissue states: (1) histopathology to ex vivo fixed MRI; (2) the result of 1 to ex vivo fresh MRI; (3) the result of step 2 to in vivo MRI.…”