Computer simulation offers the possibility to train, plan, and predict medical interventions on a computer system. A computer model of the anatomy of interest is developed, and clinicians can interact with this model without any risk to patients. From a technical point of view, medical simulation requires addressing many of the challenges researched in our group, in particular deformation modeling and contact handling. In addition, medical simulation poses additional challenges such as real-time computational requirements and patient-specific behavior.
In this line of research we collaborate with GMV.
Deformation of Medical Images
Many inherently deformable structures, such as human anatomy, are highly heterogeneous. We develop methods to interactively deform heterogeneous volumes, which can be used for medical training and planning applications based on CT and MRI images. This problem involves several challenges on both deformation and visualization of regular volumetric discretizations.
Our approach rests on two major components: a massively parallel algorithm for the rasterization of tetrahedral meshes (Gascon et al. 2013) and a method to define homogenization, which is the process of determining parameters of a coarse discretization that best matches the behavior of the fine discretization (Torres et al. 2014).
In (Gascon et al. 2013
) we perform rasterization as a massively parallel operation on target voxels, and we minimize the number of voxels to be handled using a multi-resolution culling approach. Our method allows the deformation of volume data with over 20 million voxels at interactive rates.
In (Torres et al. 2014
) presents a method to interactively deform volume images with heterogeneous structural content, using coarse tetrahedral meshes. It rests on two major components: the previous massively parallel algorithm for the rasterization of tetrahedral meshes, and a method to define a coarse deformable tetrahedral mesh from the homogenization of a fine heterogeneous mesh. We show the potential of the method for training and planning applications through two examples: an abdominal CT exploration and the alignment of breast CT and MRIs.
Human joints, such as the shoulder, present intricate connections of anatomical elements such as bones, muscles, tendons, ligaments, and fat. The nature and arrangement of the various structures in the shoulder impose two main difficulties for interactive simulation: a large diversity of mechanical properties, ranging from hard bone to soft fat tissue, and complex contact situations. In (Otaduy et al. 2010)
, we present a combination of representations, simulation methodology, and algorithms, which, altogether, provide the proper balance between simulation quality and performance for interactive medical applications. Unified representations for all dynamic objects and their dynamic state allow us to define coupling constraints and contact constraints in a general way. As a result, all dynamic objects can be simulated at once in a unified manner. We show the application of our algorithm to shoulder simulation in two medical settings: virtual arthroscopy and physiotherapy palpation.