Patient-specific organ-scale prediction of prostate cancer growth supported by imaging-based immersed isogeometric analysis
Please login to view abstract download link
Prostate cancer (PCa) remains one of the most common and lethal malignancies in men worldwide. Despite advances in early detection and treatment, the limited personalization and the observational strategies that characterize PCa clinical management make over- and undertreatment two key unresolved issues. To address these limitations, we propose a personalized forecasting framework based on biomechanistic computer simulations supported by isogeometric analysis [1]. Focusing on active surveillance (i.e., monitoring of common, indolent disease after diagnosis and before treatment), our approach reconstructs the patient-specific prostate geometry from T2-weighted MRI data and personalizes the parameters controlling tumor dynamics via a nonlinear least-squares calibration to longitudinal multiparametric MRI data. Our model further enables prediction of serum prostate specific antigen (PSA) levels using longitudinal measurements of this common biomarker of PCa burden. We utilize the finite cell method [2] in our isogeometric approach. This immersed boundary framework defines the MRI-informed prostate geometry as a level-set function, embedding it within a hexahedral tensor-product mesh aligned with the T2-weighted MRI voxel grid. This approach circumvents the need of complex boundary-fitted mesh generation, enabling the direct use of MRI data and segmentations within a simple computational domain. Additionally, it facilitates local refinement around the tumor using a truncated hierarchical spline basis [2]. Our results demonstrate that our computational approach effectively captures individualized PCa dynamics, providing a robust and efficient platform for personalized tumor forecasting. We posit that clinical validation of this technology could support more informed, patient-specific clinical decision-making and help combat over- and undertreatment of PCa. REFERENCES [1] G. Lorenzo, et al. (2024). A pilot study on patient-specific computational forecasting of prostate cancer growth during active surveillance using an imaging-informed biomechanistic model. Cancer Research Communications, 4(3), 617-633. [2] M.J. Johnson et al (2023). Image-guided subject-specific modeling of glymphatic transport and amyloid deposition. Computer Methods in Applied Mechanics and Engineering, 417, 116449.
