3D visualization of the joints using MR images.
Most images are 2D in nature and thus surgeons have to use multiple images (aligning or comparing them) in order to construct a 3D model from their imagination for surgery. Traditionally, a radiologist supposed to diagnose such data sets would look at the scanned images slice by slice. By mentally r...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
2009
|
Subjects: | |
Online Access: | http://hdl.handle.net/10356/16401 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | Most images are 2D in nature and thus surgeons have to use multiple images (aligning or comparing them) in order to construct a 3D model from their imagination for surgery. Traditionally, a radiologist supposed to diagnose such data sets would look at the scanned images slice by slice. By mentally reconstructing the sampled information into a three-dimensional representation, he would judge on the health of the patient. (König et al., 2001) However, we know that in science, and more specifically in the application of medicine, we need to quantify the methods and having a qualitative analysis is definitely insufficient.
Computer assisted surgery have become increasingly important in pre-operative planning and during operation. This involves the use of patient specific 3D computer anatomical models, generated using 3D imaging modalities (e.g. CT, MRI). These models
are approximated from segmented CT scans of the bones. (Grimm et al.) Adequate visualization (e.g. bone, cartilage and ligaments) using 3D models are important in allowing accurate and clinically meaningful assessment of knee joint for surgery. The aim of this project is to develop visualization techniques that can assist surgeons to plan and perform the surgery.
Using a visualization pipeline, there was an achievement of the ultimate goal of 3D volumetric model reconstruction of the knee joint from 2D segmented MRI images. This allows surgeons to develop a better feel of what they will operate on and prepare them more adequately for surgery. They can also perform precise measurements on the model.
Additional interactive functions were also added so that surgeons may undertake tasks such as assessing the model, point-picking, and varying the visibility of the structures so that the region of interest can be selected while disregarding the surrounding regions.
The segmentation of a few structures was done - the tibia and the patellar. The segmented images were further processed to reconstruct the final model, consisting of 3D visualization and final 3D model reconstruction of the various structures, integrated to form a full working knee joint model. An algorithm was developed to take in different user inputs as directories for the segmented images and automatically assemble the final model from these source images. Last but not least, the significance of this project, visualization of the knee joint is portrayed by the visualization of inter-bone distances that has the potential to characterize 3D structures and spatial relations non-invasively in complex joints. (Demiralp et al., 2001) |
---|