Toward a real time multi-tissue Adaptive Physics-Based Non-Rigid Registration framework for brain tumor resection

Our recent work to improve the accuracy of non-rigid registration while maintaining its completion within the neurosurgery time constrains published in Frontiers in Neuroinform., 17 February 2014. The new method can reduce the alignment error up to seven and five times compared to a rigid and ITK's physic-based non-rigid registration methods, respectively. On average, the alignment error of the new method is reduced by 9.23 and 5.63 mm compared to the alignment error from the rigid and PBNRR method implemented in ITK.

Medical Image Non-Rigid Registration (NRR)


Image Guided Neurosurgery (IGNS) is different from conventional neurosurgery by the availability of imaging which helps surgeon to understand what is going on inside the patient's brain while the tissue resection is underway. The major advantage of this is that the surgeon can avoid the critical regions of the brain, as they shift due to changing conditions as the surgery goes on. Thus, there is a better possibility for good outcome of the operation and improved quality of his or her life afterwards.

Non-rigid registration is one of the enabling technologies for IGNS. As the surgery progresses, the brain changes its shape and shifts. Thus, multimodal imaging data acquired prior to the surgery becomes invalid. During IGNS brain deformation is tracked by acquiring low resolution MRI in an open MR scanner, which is subsequently used to deform, or register, the preoperative data.

Preop Intraop
Preoperative MRI Intraoperative MRI


Volumetric registration of MRI using intraoperative imaging has long been too computationally expensive to be practical. Our distributed implementation of non-rigid registration, for the first time, enabled in-time delivery of registration results to the surgeons. The implementation we have developed is able to span multiple computational sites distributed geographically, and communicate the results back to to operating room.

Neurosurgery 1 Deformation overlay
Deformation field is computed using intraoperative image
and mesh-based biomechanical model.

Image to Mesh Conversion


The goal of this project is to create fast and robust software for generating tetrahedral meshes from (segmented) image data. This task challenges traditional approaches to mesh generation, as no surface information is available as such.

We explore two conceptual approaches to mesh generation: (1) segmentation of the structure of interest with subseqent surface recovery and applying state of the art methods to tesselate the volume, and (2) generate the volume mesh directly from the segmentation (or even unprocessed image).


Currently, there are two main driving applications for this project:image-guided neurosurgery(IGNS), and large scale simulation of arterial blood flow.

The mesh generator we developed facilitates construction of biomechanical model of brain for intraoperative non-rigid registration of preoperative MRI. The mesher works directly with the segmented image data, it is fast and generates meshes of quality comparable with the state of the art counterparts, which require object surface to be recovered first.

Up until recently, our implementation was enabling intraoperative registration studies at Brigham and Women's Hospital, Boston.

Brain mesh surface Graded mesh cut
Vol mesh with bg Surface and bg

Tetrahedral mesh is essential for construction of biomechanical model of brain deformation.

Our collaboration project targeting construction of tetrahedral meshes for arterial blood flow simulations is currently in the early stages of development. Our initial objectives of the study in this area is the assessment of available approaches and evaluation of application-defined requirements.

Artery hires Bifurcations

Surface of the tetrahedral mesh recostructed from segmented artery volume;
color-coded is volume element quality as measured by aspect ratio.


The following software is used in this project.

Related Links