ePrints.FRI - University of Ljubljana, Faculty of Computer and Information Science

An Analysis of Cell Nucleus Images in Mammary Gland Tissue Using Computer Vision

Katarina Mele (2000) An Analysis of Cell Nucleus Images in Mammary Gland Tissue Using Computer Vision. Prešeren awards for students.

Full text not available from this repository.


The thesis combines the subject areas of computer vision and medicine. It deals with cell nucleus segmentation and the boundary detection of the groups formed by nuclei. The fact is that it is both cell nucleus features and group shapes that serve as cancer diagnosis criteria. Also further treatment is dependent on the information about the cell nucleus spatial arrangement and the reoccurrence of the shapes formed by malignant, potentially malignant and normal ducts. In addition, the work includes several procedures of cell nucleus segmentation described in previous work. The author has developed a method for the boundary detection of nucleus clumps. It is based on grouping techniques with the greedy algorithm, relaxation and the graph search. The analysis is followed by some examples showing the use of the boundary detection method. Differences between normal and malignant ducts are shown on the basis of some images of marnrnary glands. Finally, some improvements and future work extensions are suggested. This is a new approach in the area ofcitometrics and DCIS, since it is a well-known fact that estimates of the architectural characteristics are hardly recurrent. If it were possible to make an estimate objective by means of the duct architecture processing programme, more reliable and recurrent information about the tumor would be obtained.

Item Type: Thesis (Prešeren awards for students)
Keywords: computer vision, sgementation, relaxation, graph search, greedy algorithm, breast cancer, cytology
Number of Pages: 89
Language of Content: Slovenian
Mentor / Comentors:
Name and SurnameIDFunction
prof. dr. Aleš Leonardis29Mentor
Link to COBISS: http://www.cobiss.si/scripts/cobiss?command=search&base=51012&select=(ID=3483220)
Institution: University of Ljubljana
Department: Faculty of Computer and Information Science
Item ID: 3732
Date Deposited: 06 Jan 2017 11:01
Last Modified: 13 Feb 2017 11:31
URI: http://eprints.fri.uni-lj.si/id/eprint/3732

Actions (login required)

View Item