Novel analytical imaging techniques like MRI, functional MRI, Diffusion Tensor Imaging (DTI), Computer Tomography (CT), Positron Emission Tomography (PET) have improved medical diagnosis. However, the information from these advanced imaging techniques has not been fully exploited for the systematic design of effective drug administration. Effective drug diffusivity, hydraulic soft-tissue properties, metabolic reaction constants and binding coefficients are often determined in simplified bench-scale experiments. There appears to be a gap between high quality of imaging techniques and their use in quantitative analysis in the clinical practice. The sketchy understanding of the intracranial dynamics and drug transport mechanisms hampers the clinical introduction of invasive drug administration for the treatment of neurodegenerative diseases.

Proposed approach. We propose to integrate state-of-the-art imaging techniques and first principles models for transport phenomena to design systematically the best delivery option for each disease or drug. Our hierarchical design procedure features three innovations:

A. Accurate three-dimensional reconstruction of the patient-specific brain geometry

B. Prediction of achievable volumes of distribution in anisotropic and heterogeneous brain tissues

C. Rigorous mathematical programming for optimal catheter design and placement

A. Physiologically consistent brain reconstruction. We accurately reconstruct the three-dimensional patient-specific brain geometry and specific substructures of the midbrain close to potential drug infusion sites. We use state-of-the-art geometric image reconstruction tools to construct computational grids from novel MRI or histological data.

B. Prediction of drug distribution. We calculate achievable volumes of distribution in anisotropic and heterogeneous regions of the brain from in-vivo image data. We create a three-dimensional map of the brain using diffusion tensor imaging (DTI). We postulate that the effective drug transport is correlated to the fiber alignment. By accurately accounting for the anisotropy and heterogeneity of the soft brain tissue, we expect to substantially improve the fidelity of drug distribution predictions. Moreover, we also exploit DTI data to assess the hydraulic permeability of the brain tissue. We have developed novel mathematical inversion techniques allowing us to accurately ascertain unknown metabolic reaction rates like drug formation or clearance, receptor binding, or accurately determine transport properties of the brain tissue such as drug diffusion tensor, the permeability tensor, porosity or fluid friction coefficients.

C. Optimal catheter design and placement. We address the third challenge of identifying the optimal delivery parameters. Specifically, we find with rigorous optimization techniques the optimal placement and orientation of the infusion catheter, its dimensions, shape, and number of drug release ports specific to the target area and the bimolecular properties of the drug. In addition, the effect of infusate backflow along the catheter shaft prevalent at high infusion rates will also be addressed.

Impact. We expect that the computer-aided method will create new knowledge about drug distribution in the brain. It will provide a scientific tool for better interpreting existing experimental in vivo animal or human data.