A novel tool for directionality assessment in medical images
Danilova, Maria (2019)
2019
Electrical Engineering
Lääketieteen ja terveysteknologian tiedekunta - Faculty of Medicine and Health Technology
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2019-05-23
Julkaisun pysyvä osoite on
http://urn.fi/URN:NBN:fi:tty-201905211723
http://urn.fi/URN:NBN:fi:tty-201905211723
Tiivistelmä
Modern medicine relies on visual data. Until recently, manual inspection was the only method available, although it introduced a large amount of subjectiveness into the final diagnosis. Computer-aided medical data processing aims to increase diagnosis accuracy. This thesis covers one method to use computers for determining image directionality, which can be applied in nerve and vessel disease detection.
In particular, a number of directionality evaluation methods for 2-dimensional (2D) images are examined. For one of the methods, a 3-dimensional (3D) extension is proposed. This method is based on the Fourier Transform. The relationship between the spacial domain lines and their frequency domain counterparts and the way it can be used to define the directionality is covered in the ‘Materials and methods’ part of this thesis.
The proposed extension was implemented using MATLAB and supplied with a graphical user interface (GUI). It was tested against both test images, such as lines and volumetric spheres, and real optic nerve MicroCT images. Results of these tests were analyzed and suggestions were given on the potential usage of the method.
In particular, a number of directionality evaluation methods for 2-dimensional (2D) images are examined. For one of the methods, a 3-dimensional (3D) extension is proposed. This method is based on the Fourier Transform. The relationship between the spacial domain lines and their frequency domain counterparts and the way it can be used to define the directionality is covered in the ‘Materials and methods’ part of this thesis.
The proposed extension was implemented using MATLAB and supplied with a graphical user interface (GUI). It was tested against both test images, such as lines and volumetric spheres, and real optic nerve MicroCT images. Results of these tests were analyzed and suggestions were given on the potential usage of the method.