 |
07/1 Acquisition, reconstruction & interactive visualisation
Generic functional model of joints (consisting of bones, cartilage, ligaments, and muscles) that can be acquisition, reconstruction & interactive visualisation
Generic functional model of joints (consisting of bones, cartilage, ligaments,
and muscles) that can be partly individualised will be reconstructed. Relevant
patient's anatomical and kinematical data extracted from images (MRI, CT) and
motion capture (dynamic MRI, optical motion capture), as well as statistical
data, will be used to adjust the generic model to clinically relevant feature
of the patient's joint. The main challenge is the extension and completion of
the existing acquisition and reconstruction techniques to establish a
clinically applicable examination framework.
Anatomical reconstruction
Generic anatomical models of bones and soft-tissues (cartilages, ligaments,
and muscles) of the hip and knee joints and topological interrelationship
between anatomical components of two volunteers will be constructed
interactively from MRI and/or CT data. Automatic methods for the template-based
reconstruction of patients' bones based on deformable models developed for the
femur and the pelvis will be generalised. The adjustment of generic
soft-tissues models with some clinically relevant individual features
(ligaments/ muscles attachment points and action lines, cartilage thickness)
will be investigated. Image processing and inter-patient registration methods
will be developed for this purpose. In addition, statistical cartilage
thickness data will be used in the individualisation process.
Kinematical reconstruction
The kinematical model developed in the 1st phase (data and conception) consists
of a set of motion/deformation information defining the behaviour of an
articulated system. This model was limited to healthy and simplified hip joints
(bones and skin). The kinematical reconstruction methods from medical data will
be generalized to include all functional organs, and some pathological cases.
Novel methods for motion tracking in dynamic MRI will be investigated in order
to improve computational performances and the accuracy of the tracking with
regards to acquisition time and spatial resolution. This can be also applied
for segmentation (i.e. anatomical reconstruction") and registration (i.e.
morphological organs comparison in "clinical applications").
Interactive visualisation
Based on the spreadsheet interface developed in 1st phase, a clinical
examination framework with multimodal medical data inputs will be developed
with the integration of different modules (modeling and simulation modules) for
the interactive visualisation of the hip joint in motion. This multimodal
data-driven framework will allow the visualisation of individualised body
motion by merging several types of data acquired from different kinds of
acquisition techniques (MRI, optical motion capture, body scan, EMG). Moreover,
in case of sparse input data sets (as in most of clinical sites only a few of
the acquisition equipments will be available), missing information will be
estimated from generic models. This approach defines a level of details of
accuracy depending on the range of data that can be acquired from a given
patient. This framework will be developed with regards to physicians
requirements (c.f. sub-project "medical consulting").
Last update of project infos on 2009-05-19.
|
