Thesis /

ROMDOC-THESIS-2017-1024

Contribuţii la analiza, testarea şi predicţia dezvoltării tumorilor canceroase apelând la tehnici inteligente

Bălănică, Victor
2011-10-07

Abstract: This thesis aims to develop a computer platform that allows computer-aided diagnosis of cancer diseases in the particular case of breast cancer through analysis, classification and prediction techniques of the development of cancerous tumors applied on mammographic imaging results. In the analysis of the tumoral mass, the spiculation degree of the tumoral edge is a particularly relevant indicator of malignancy, which allows distinguishing malignant from benign lesions. The thesis introduces four new methods for the extraction and assessment of the tumor contour on a number of boundary neighborhoods: Method 1 - The Maximum Level Difference for every region, Method 2 - The Total Area of Triangles for every neighborhood, Method 3 - The Total Angle for every region and Method 4 - The Total Quadratic Curvature for each region. The resulting feature sets and also other tumor characteristics available in the database are used in training a neural classifier and a more efficient hybrid geno-neural classifier designed for the evaluation and differentiation of the tumors identified on the mammograms of the new cases. Subsequently, based on the fuzzy set theory, a number of techniques are developed for analyzing the malignancy on the BIRADS assessment scale, for diagnosing malignancy (benign / malignant) and also a geno-fuzzy method for the same purpose but with better performance. The evaluation of the development direction of the tumor and the assessment of the treatment progress is achieved by a new method based on the computation of the Euclidean, Hamming and Leventstein metrics of the mammographic results. Were proposed two CAD prototypes modules that implement intelligent techniques, trained using old cases information and data and useful in predicting and recommending optimal treatment for a new examined case of cancer. Finally, the implemented software platform, MAMMA, for the management, the CAD analysis and the decision support of the diagnosis regarding the classification and the prediction of the cancerous tumors growth is presented.

Keyword(s): Imagistică medicală -- Tumori -- Teză de doctorat ; Cancer de sân -- Diagnoză -- Teză de doctorat ; Prelucrarea imaginii -- Tehnici digitale -- Teză de doctorat ; Diagnoză medicală -- Tehnici digitale -- Teză de doctorat
OPAC: See record in BC-UPB Web OPAC
Full Text: see files

Similar records

 

People who viewed this page also viewed:

(261)  Optimizarea conceptuală şi operaţională a instalaţiilor chimice multiscop - Voinescu, Sorin - ROMDOC-BC_UPB-THESIS-2003-000000054 (255)  Tehnologiile informării şi comunicării : suport de curs - Curta, Olimpia - ROMDOC-BOOK-2007-005 (252)  Spectrometrie de masă cu acceleratorul în zona nucleelor actinide - Păceşilă, Doru Gheorghe - ROMDOC-THESIS-2020-2165 (250)  LinkContribuţii la cercetarea compozitelor in-situ Al-Zn-Mg-Cu/TiB₂ produse prin reacţie aluminotermică - Marcu, Dragoş Florin - ROMDOC-THESIS-2020-2141 (248)  Scientific Bulletin Series C:Electrical engineering - ROMDOC-Journal-2007-003

 

Rate this document:

Be the first to review this document.

Discuss this document:

Start a discussion about any aspect of this document.