| Author: | Poon, Yiu-man |
| Title: | Intelligent landmark-assisted CT image reformation aligning with the neuro-ocular plane and its perpendicular |
| Degree: | M.Sc. |
| Year: | 2010 |
| Subject: | Hong Kong Polytechnic University -- Dissertations Tomography Diagnostic imaging -- Computer simulation |
| Department: | Department of Health Technology and Informatics |
| Pages: | x, 102 leaves : ill. (some col.) ; 30 cm. |
| Language: | English |
| Abstract: | Computed Tomography (CT) of orbits could provide up to 300 images in just a few seconds of scan time. In order to produce thicker images for long-term archive, electronic patient record (ePR) and filming, radiographer should additionally prepare multiplanar reformation (MPR) manually which usually requires additional 5 minutes. Two sets of reformatted images are axial and coronal. They should be parallel and perpendicular to neuro-ocular plane (NOP). The NOP is defined as plane passing through the lenses, the optic nerve heads and the optic canals, with the patient maintaining primary gaze. The importance of NOP is the correspondence to the human visual pathway. For CT and MRI, axial images parallel to the pathway provide maximum information for detecting abnormality. If the MPR can be done with minimal human intervention, time can be saved. Frankfurt-Virchow plane (FVP) is one of the classic planes describing the horizontal plane of human skull. The angulation between NOP and FVP is reported as 7° from European studies. A new plane proposed is rhinion-porions plane (RPP) where three bony landmarks (porions and rhinion) are in same plane. These three landmarks are sharp for computer-aided detection (CAD). If the angulation between RPP and NOP is found and CAD for the landmarks is successful, computer-aided MPR is possible. The aim of this study is to demonstrate the feasibility of computer-aided detection of promising landmarks for MPR, in order to expedite clinical workflow. Two hundred cases of CT orbits, paranasal sinuses and facial bones were retrospectively collected from a local hospital archive. For each case, the angles of the FVP, RPP and NOP with respect to the horizontal scan plane were measured at workstation. The angles of lateral rotation and horizontal tilting of the head were also measured. Results indicates the angle between NOP and FVP is 2.9° (SD=4.2°; 95% CI=[2.3°, 3.5°]); gender difference was significant (mean difference: 1.7°; p<0.01). The angle between NOP and RPP was 4.5° (SD=4.5°; 95% CI=[3.9°, 5.1°]); gender difference was significant (mean difference: 2.0°; p<0.01). A CAD programme was written to detect porions and rhinion from CT images. The first step of the CAD was to locate the soft tissue ear openings by comparing the change of air area between consecutive images. The second step was using the location of soft tissue opening to locate the bony ear opening which was porion. After locating both locations of porions, a perpendicular bisector was constructed between porions. The third step was to detect the tip of nasal bone along the bisector. The lowest point of the tip was rhinion. The CAD was tested with these 200 cases, 47% of them were successful. The coordinates of the detected landmarks were used to calculate the angles between RPP and scan plane. Comparing the measured angles with the calculated angles, small difference was found. In summary, this study hypothetically showed that RPP can be detected and the angulation of NOP can be calculated because constant relationship between RPP and NOP is observed. This study also suggests that computer-aided MPR for image aligning and perpendicular to NOP is feasible. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| b23526543.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 18.19 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/5637