Back to search

National and Kapodistrian University of Athens

Συσχέτιση παθολογοανατομικών ευρημάτων και υπολογιστικής ανάλυσης της ψηφιακής μαστογραφίας με χρήση προηγμένων τεχνικών επεξεργασίας εικόνας

Abstract

dc:description

BACKGROUND: The aim of this study was to correlate the computed analysis of digital mammography with the pathogists results of the breast biopsy. More specifically, we analysed and studied the texture features in different region of interest of the digital mammograms that could reveal the diffent architecture of each breast lesion and additionally the potentiality of a computer aided diagnosis system to discriminate the real benign microcalcifications among a subset of patients with BIRADS 3 mammograms who underwent biopsy, thus giving radiologists the opportunity to downgrade from BIRADS 3 to BIRADS 2 category in certain cases and thereby eliminate unnecessary biopsies in probably benign lesions. METHODS: We analyzed 109 consecutive BIRADS 3 mammograms of patients that underwent Stereotactic Vaccum Assisted Breast Biopsy aand compared them to the pathologist results. The computer aided diagnosis (CAD) system Hippocrates-MST detected and quantified critical features of microcalcifications and classified them according to their probability of malignancy on a risk percentage scale. In addition we extracted the texture features (Analysis) of non palpable mammographic lesions of 60 women who underwent vaccum assisted breast biopsy (MammotomeR) on the stereotactic table of Fischer’s in every digital image that was acquired during the biopsy. The features that were evaluated were 1st order statistic features ( minimal, maximal, median, and mean gray levels, standard deviation of gray levels, coefficient of variation and gray level skewness, kurtosis, energy and entropy) and fractal dimension as well. According to the pathologist reports, the set was comprised from 13 DCIS, 16 IVC, 10 precancerous (LN, ADH) lesion and 21 benignities. RESULTS: The system Hippocrates MST proposed biopsy for all cancer cases. Unfortunately it missed 4 atypical lesions which should be referred for surgical excision because of their high probability of having or developing malignancy, causing the system’s sensitivity to decrease to 81% with 95% CI (58–94). However, the system downgraded 29 out of88 benign calcification clusters, previously assigned to BIRADS 3, to BIRADS 2 category, and so achieved a reduction of 33% inunnecessary biopsies. On the other hand, the statistical analysis of the results of Analysis system (Mann-Whitney U test) revealed significant statistical difference between the IVC and the DCIS in minimal gray level (p=0.0028), in standard deviation (p=0.009), in entropy (p=0.03) and in energy (p=0.018). There was statistical difference in the energy and the standard deviation between IVC and precancerous lesion and also in the entropy between IVC and benignities. Finally, the the data set was statistically analysed with Kruskal-wallis test the relationship among the median values of the breast lesions and there was significant statistical difference in standard deviation (p=0.0488) and energy (p=0.0476). In addition, we calculated a range of AUC, 0.6949-0.7885 for the characteristic that were statistically important. The higher AUC was the one for standard deviation when comparing the DCIS vs IVC. CONCLUSION: The results justified that the CAD system Hippokrates MST has the potential to improve radiologist’s performance in the detection and evaluation of microcalcifications, as it revealed the malignant clusters among BIRADS 3 mammograms and unveiled 33% of the real benign cases, consequently downgrading them from BIRADS 3 to BIRADS 2category. Therefore, with further development and investigation using larger samples it could be used to detect cancerous lesions and eliminate unnecessary biopsies of benign lesions. In addition, the computed texture analysis of mammograms with Analysis system and their correlation to the pathologist results has important results in differentiation of breast pathology and could become an important ally for the radiologist and the physician in early diagnosis and therapy of breast pathology.

Degree

thesis:*
Grantor dc:publisher
National and Kapodistrian University of Athens
Year dc:date
2013

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Giannakopoulou, Georgia
  • Γιαννακοπούλου, Γεωργία

Subjects

dc:subject × 29

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/39803

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Giannakopoulou, Georgia; Γιαννακοπούλου, Γεωργία. Συσχέτιση παθολογοανατομικών ευρημάτων και υπολογιστικής ανάλυσης της ψηφιακής μαστογραφίας με χρήση προηγμένων τεχνικών επεξεργασίας εικόνας. National and Kapodistrian University of Athens, 2013. http://hdl.handle.net/10442/hedi/39803