Masters Research - Master of Medical Science / Introduction: Between 6-10% of women attending breast screening are recalled to investigate an unclear area on the mammogram. Image quality is known to affect image interpretation and it has been suggested that the number of recalls could be reduced with improved image quality. Aim: This study aimed to investigate the effect image quality has on recall rates, to assess reader consistency using the PGMI classification system and to establish factors leading to recall. Materials and Methods: A six member panel assessed 904 sets of images (698 recalled; 206 non-recalled) through a BreastScreening Program during three separate phases (pilot, main and non-recall). The pilot study was conducted without additional training in PGMI. Levels of agreement and Kappa statistics were calculated to assess intra- and inter-consistency. The percentage of and reasons for inadequate images was calculated; while a case-control study was conducted to establish factors increasing the likelihood of a client being recalled. Results: The level of agreement between panel members significantly increased from the pilot to the main study (45.5% to 57.7%) before decreasing slightly for the non-recall (57.7% to 52.2%). Overall, 3.3% of the 904 sets of images were classed as inadequate; the most common PGMI reason was exposure (31%); the left MLO was considered the most common inadequate projection (30%), with more privately produced (66%) images considered inadequate compared to public images (34%). Inadequate image quality did not hinder the cancer detection rates. The case-control component demonstrated current and previous HRT use, increased breast density, better image quality and images being taken at a public site all contributed to a client being recalled. Conclusion: The results of this study demonstrated that inadequate image quality was not a major factor leading to recall; although twice the number of recalled images were considered inadequate compared to the non-recalled images. The use of the PGMI classification system is highly subjective, with low levels of agreement amongst users. The use of HRT, breast density, imaging site and image quality all contribute to a client being recalled.

The mean glandular dose (MGD) was measured for a breast phantom by using molybdenum/molybdenum and molybdenum/rhodium target/filter combinations, at different kVp 26, 28 and 32 kilovolts. The phantom thickness was 7.5cm and was made of BR12 material. The change of dose was studied as a function of depth inside the phantom at different depths from the surface, namely 3.3, 4.3 and 5.3cm, by using TLDs. It was found that the MGD value for different combinations of beam quality (HVL) and energy (kVp) did not exceed the recommended values given by different protocols. The Mo/Rh target/filter required lower doses to achieve the same or better results compared with the Mo/Mo target/filter. The change in the surface dose as a function of kVp was more significant for Mo/Rh than for the Mo/Mo. Studying the change in dose within the breast, as a function of depth gives a better understanding of the interactions between radiation and tissue inside the breast. It should be noted that the MGD is a tool for optimization of the mammography parameters. However, the MGD should not be used directly to estimate the risk of determinable health effects from mammography. This will ultimately help to determine limits for the breast surface dose and a better understanding of cancer risk. In future work, we will try to measure the change of the dose as a function of depth by using more kVp, HVL, different breast composition and different target/filter combinations to give a wider picture for different situations.

Sequential mammograms contain important information, such as changes of the breast or developments of the masses, for diagnosis of disease. Comparison of sequential mammograms plays an important part for radiologists in identifying malignant masses. However, currently computer-aided detection (CAD) programs can not use such information eciently. The diculties lie in the registration of sequential mammograms. Most of current methods register sequential mammograms based on control points and image transformations. For these methods to work, extraction and correspondence of the control points is essential. This thesis presents a new approach in registering mammograms. The proposed method registers mammograms by associating mass-like objects in sequential mammograms directly. The mass-like objects appear in the images of normal breasts as well as images of breast with cancer. When the mass-like objects in sequential mammograms are accurately associated, measurements of changes in mass-like objects over time become possible. This is an important way to distinguish mass-like objects associated with cancer from cysts or other benign objects. The proposed method is based on graph matching. It uses the internal structure of the breast represented by the spatial relation between the mass-like objects to establish a correspondence between the sequential mammograms. In this method, the mammogram is firstly segmented into separate components using an adaptive pyramid (AP) segmentation algorithm. A series of filters, based on the features of components, is then applied to the components to remove the undesired ones. The remaining components, the mass-like objects, are represented by a complete graph. The spatial relations between the remaining mass-like objects are expressed by fuzzy spatial relation representation and are associated to the edges of the graph as weights. Association of the mass-like objects of two sequential mammograms is realized by finding a common subgraph of the corresponding two graphs using the backtrack algorithm. The segmentation methods developed in the course of this work were tested on a separate problem in computer-aided detection of breast cancer, namely the automatic extraction of the pectoral muscle. The graph matching method was tested independently of the segmentation method on artificially distorted mammograms and the full process, including the segmentation and the graph matching, was evaluated on 95 temporal mammogram pairs. The present implementation indicates only a small improvement in cancer detection rates but also presents opportunities for a substantial development of the basic method in the future.

Diss. Umeå : Umeå universitet, 2008.

- / -

Orientador: Prof. Dr. Harki Tanaka / Dissertação (mestrado) - Universidade Federal do ABC, Programa de Pós-Graduação em Engenharia Biomédica, 2015. / Esse trabalho visou realizar um estudo para auxiliar gestores de tecnologias médicas por meio da denição de critérios de avaliação de tecnologia em saúde na escolha de um equipamento de mamograa. O gestor de tecnologias médicas na atualidade encontra no mercado três tipos de tecnologias de mamograa diferentes: mamograa convencional, mamograa computadorizada e mamograa digital. Primeiramente foi realizado um estudo sobre a cadeia de imagem de cada tecnologia. A partir do conhecimento da cadeia de imagem de cada tecnologia, foi possível identicar cada processo característico de todas as tecnologias e com isso identicar a especicidade do custo gerado por essas modalidades em cada ponto. Alguns valores como custo de aquisição, custo de manutenção e custo de insumos foram coletados da Bolsa Eletrônica de Compras do Estado de São Paulo e no Diário Ocial da União. Características relacionadas à epidemiologia, sensibilidade e especicidade foram coletadas em estudos cientícos. Um estudo sobre a desinstalação e descarte foi realizado, levando-se em consideração as diretrizes para cada tecnologia. A m de nos aproximar da realidade do funcionamento de um mamógrafo, as diferentes tecnologias em mamograa foram inseridas em um contexto de uma clínica de imagens radiológicas. Para auxiliar a análise econômica foram desenvolvidas planilhas de cálculo eletrônico. A vida útil considerada para cada equipamento foi de 10 anos. O custo que representou maior impacto econômico foi o de recursos humanos em todas as tecnologias. A mamograa que apresentou o menor custo por exame foi a mamograa computadorizada sem impressão do lme (CRP), embora o custo de aquisição dos seus equipamentos seja aproximadamente quatro vezes maior que a mamograa convencional. A mamograa computadorizada quando existe a impressão do lme apresentou o maior valor por exame. Os estudos de acurácia e sensibilidade mostraram que a mamograa digital tem um pequeno ganho apenas em mulheres abaixo de 50 anos quando comparada à mamograa convencional. / This study aimed to help medical technology managers to choose mammography equipment through criteria of health technology assessment. The manager of medical technologies nds on the market today three dierent types of mammography technologies: conventional mammography, computerized mammography and digital mammography. First, we conducted a study on the image chain for each technology. From the knowledge of the image chain for each technology, it was possible to identify each characteristic process of all technologies and thereby identifying the specicity of the costs generated by such arrangements on each point. Some values as acquisition cost, maintenance cost and cost of inputs were collected from Electronic Trading Platforms of São Paulo state and in the Ocial Gazette of Brazil. Characteristics related to epidemiology, sensitivity and specicity were collected in scientic studies. A study on uninstalling and disposal was carried out, taking into account the guidelines for each technology. In order to approach the reality of the operation of a mammography unit, the dierent technologies in mammography were inserted in a context of clinical radiological images. To assist the economic analysis, electronic spreadsheets were developed. The lifetime considered for each equipment was 10 years. The greater economic impact was the cost of human resources in all technologies. A mammogram that showed the lowest cost per exam was computerized mammography without screen-lm (CRP), although the cost of purchasing of the equipment is approximately four times higher than conventional mammography. Computerized mammography, when there is the impression of the lm, has the highest cost per exam. The accuracy and sensitivity studies have shown that digital mammography has a small gain only in women under 50 years of age, when compared to conventional mammography.

The South African national calibration facility is currently not equipped for mammography, dose-measuring equipment. A therapy X-ray machine was used as a calibration unit at the national secondary standards dosimetry laboratory (SSDL) for medium and low energy X-ray, therapy calibrations. It is not necessarily intuitive that the latter calibrations are applicable to diagnostic X-ray beams generated by high frequency generators. The response of measuring equipment calibrated in a therapy X-ray beam, compared to its response in a diagnostic or clinical mammography unit, is unknown. The aim of the research was to investigate whether there was a measurable difference between the X-ray beam qualities available for low energy diagnostic radiology and radiation therapy, i.e. up to 100 kV. The beam qualities studied included both mammography and conventional diagnostic radiography, i.e. nominally 20 kV to 100 kV. The diagnostic and therapy X-ray tubes under investigation had different target-filter combinations, inherent filtration and theoretically, different X-ray spectra. Practically, spectrometry of X-ray beams is not possible because of the sophistication of the instrumentation, comprehensive analyses being very time consuming and not practically applicable to the clinical environment (Kharatti and Zarrad, 2003). Furthermore, not all SSDL’s or Hospitals have access to spectral analysers. Clinical beam quality is instead specified in terms of both the tube peak voltage and the half-value layer (HVL), the thickness of material that will reduce the maximum output of the X-ray beam to 50%. The goal was to compare measured HVLl’s to the ones recommended by the International Electro technical Commission (IEC-61267, 2005) for available mammography beam qualities. The method was validated using attenuation curves. The attenuation curves were then used to derive the suitability of the X-ray spectra for calibration of mammography ionisation chambers (Waggener and Blough, 1999). One of the low energy therapy units was found to be suitable for introducing a regional calibration service for mammography.

This dissertation contributes to the field of health economics, which, in the past couple of decades, has substantially increased our understanding of the determinants of human health, health-related behavior, and health care choices. A large body of literature has documented the influence of peer group behavior on individual choices. The purpose of my research is to examine the extent of such a phenomenon in breast cancer preventive behavior. Using Behavioral Risk Factors Surveillance System (BRFSS) surveys from 1993-2008, I measured the effect of other female screening behavior on an individual's decision to have a routine breast cancer screening by calculating the size of a so called social multiplier in mammography. I estimated a vector of social multipliers in the use of annual mammograms by taking the ratio of group-level effects of exogenous explanatory variables to individual-level effects of the same variables. Peer groups are defined as same-aged women living in the same geographical area: county or state. Several econometric methods were used to analyze the effect of social interactions on decision to undergo mammography, including ordinary least squares, fixed effects, the split sample instrumental variable approach, and a falsification test. My results supported the hypothesis that social interactions have an impact on the decision to have a mammogram. For all women over age 40, I found strong evidence of social interactions being associated with individual's education and ethnicity. In addition, the decision for women ages 40-49 to have a screening was subject to peer influence through their place of employment and ownership of health insurance. Finally, for women age 75 and older, being married and aging were the most important channels through which peer group influenced the decision to have a mammogram. This research has important policy implications in the presence of current health care reform that reimburses breast cancer screening at 100%, while rates of mammography receipt remain below the policy goal. Furthermore, I examined the effect of the 2009 United States Preventive Services Task Force change in screening recommendations on screening behavior. I demonstrated an immediate reduction in the receipt of mammography among women of all age groups following the revision of screening guidelines. I found that in 2010, the twelve month mammography receipt decreased by 1.97 (women ages 40-49), 2.20 (ages 50-74), and 3.61 (age 75 and older) percentage points, and the twenty-four months mammography receipt decreased by 1.47 (women ages 40-49), 1.05 (ages 50-74), and 1.92 (age 75 and older) percentage points. Analysis using a two-year follow up period after the revision of screening recommendations provided further support to this conclusion.

Although there has been an increasing incidence of breast cancer among Non-English speaking background (NESB) women in many developed countries, existing screening services are being underused by these women. Studies show that the barriers to the accessibility of breast cancer screening by NESB women include their lack of awareness, low level of education, low self-efficacy and lack of social interaction with other women. This study aimed to investigate the knowledge relating to breast cancer and mammography, self-efficacy and barriers to mammography use among NESB women in an Australian regional city before and after their attendance at a health education session. This health education session aimed to increase the awareness and use of mammography among these NESB women. Two widely used behaviour theories, ‘Health belief model’ and ‘Social Cognitive Theory,’ were applied as the theoretical framework for this study. A quasi-experimental study was conducted in which the health education session was used as an intervention. Pre-test and post-test questionnaires were completed by study participants before and after the health education session. Their knowledge of breast cancer and mammography was assessed. In addition, their self-efficacy and barriers to the use of mammography were also analysed. Results indicated that informal recruitment strategies were more effective with these NESB women. Initially 49 women were recruited. Of these, 23 women (47%) attended the health education session. As data showed tertiary educated and employed women who already had mammogram/s were more likely to attend the session. After attending the health education session, the women’s knowledge relating to breast cancer and mammography was improved and the perceived barriers to the use of mammography were reduced. During a three month follow-up period, there was no change of mammogram use by the women. However, the results showed a trend of increased intention to use the mammogram over a period of two years (41.7%) compared to six months (25.0%). Based on these results, further studies are recommended to explore the beneficial outcomes of health promotion programs targeting NESB women who are not in the workforce or have a low level of education.

Οι μικροαποτιτανώσεις είναι από τις πιο σημαντικές ενδείξεις παθήσεων του μαστού και μπορεί να αποτελέσουν πρώιμη ένδειξη καρκίνου του μαστού. Πρόκειται για εναποθέσεις ασβεστίου στο μαστό με τη διάμετρός τους να κυμαίνεται από 0.1 έως 1 mm και εμφανίζονται είτε μόνες είτε σε ομάδες. Η ακριβής τμηματοποίηση (segmentation) των μικροαποτιτανώσεων στη μαστογραφία συνεισφέρει στην εξαγωγή αξιόπιστων χαρακτηριστικών μορφολογίας, που χρησιμοποιούνται στην αυτόματη διάγνωση με υπολογιστή (Computer-aided Diagnosis, CADx). Στα πλαίσια της παρούσης διδακτορικής διατριβής προτείνεται μία νέα μέθοδος τμηματοποίησης μικροαποτιτανώσεων, η οποία αρχικά εντοπίζει σημεία του περιγράμματος αυτών. Αυτό επιτυγχάνεται με την εφαρμογή της μεθόδου ενεργών ακτίνων (Active Rays), πολικός μετασχηματισμός ενεργών περιγραμμάτων (Active Contours), σε 8 διευθύνσεις και σε δύο κλίμακες του μετασχηματισμού κυματίων (wavelet transform) με φίλτρα Β-spline. Ακολούθως, χρησιμοποιείται μέθοδος επέκτασης περιοχής (region growing) για τον ακριβή προσδιορισμό του περιγράμματος της μικροαποτιτάνωσης. Ως κριτήριο για την αύξηση της περιοχής χρησιμοποιήθηκαν τα σημεία στο περίγραμμα της μικροαποτιτάνωσης, όπως αυτά προσδιορίσθηκαν από τη μέθοδο των ενεργών ακτίνων. Επίσης, υλοποιήθηκε μέθοδος ακτινικής βάθμωσης, η οποία έχει πρόσφατα προταθεί στη βιβλιογραφία για την τμηματοποίηση μικροαποτιτανώσεων, και χρησιμοποιήθηκε για συγκριτική αξιολόγηση. Οι δύο μέθοδοι τμηματοποίησης εφαρμόστηκαν σε 149 ομάδες μικροαποτιτανώσεων, κυρίως πλειόμορφων, που αντλήθηκαν από 130 μαστογραφικές εικόνες από τη βάση DDSM (Digital Database for Screening Mammography). Η ακρίβεια τμηματοποίησης των δύο μεθόδων αξιολογήθηκε από τρεις ακτινολόγους με χρήση 5-βάθμιας κλίμακας. Η ακρίβεια τμηματοποίησης της προτεινόμενης μεθόδου βρέθηκε ίση με 3.96±0.77, 3.97±0.80 και 3.83±0.89, όπως αξιολογήθηκε από κάθε ακτινολόγο, και 2.91±0.86, 2.10±0.94 και 2.56±0.76 για την συγκρινόμενη μέθοδο. Οι διαφορές στην ακρίβεια τμηματοποίησης των δύο μεθόδων ήταν στατιστικώς σημαντικές (Wilcoxon signed-ranks test, p<0.05). Επίσης, μελετήθηκε η επίδραση των δύο μεθόδων τμηματοποίησης στην απόδοση μεθόδου αυτόματης διάγνωσης (χαρακτηρισμό) ομάδων μικροαποτιτανώσεων με υπολογιστή. Η μέθοδος αυτόματης διάγνωσης στηρίζεται σε επιβλεπόμενη ταξινόμηση προτύπων χαρακτηριστικών σχήματος ομάδας αποτιτανώσεων. Συγκεκριμένα, χρησιμοποιήθηκε ταξινομητής ελαχίστων τετραγώνων – ελάχιστης απόστασης και εξήχθησαν χαρακτηριστικά ομοιότητας και διαφοροποίησης (variability) ομάδας μικροαποτιτανώσεων, τα οποία περιγράφουν τη μορφολογία μεμονωμένων αποτιτανώσεων (εμβαδόν, μέγιστη διάμετρος, σχετική αντίθεση). Η απόδοση ταξινόμησης αποτιμήθηκε μέσω εμβαδού καμπύλης παρατηρητών (ROC). Τα χαρακτηριστικά Εμβαδού και μέγιστης Διαμέτρου επέδειξαν σημαντικά υψηλή απόδοση ταξινόμησης (Mann-Whitney U-test, p<0.05) όταν εξήχθησαν από μικροαποτιτανώσεις τμηματοποιημένες με την προτεινόμενη μέθοδο ενεργών ακτίνων (0.82±0.06 και 0.86±0.05, αντίστοιχα). Η απόδοση ταξινόμησης χαρακτηριστικών που εξήχθησαν με μέθοδο τμηματοποίησης ακτινικής βάθμωσης ήταν 0.71±0.08 και 0.75±0.08, αντίστοιχα. Συμπερασματικά, η προτεινόμενη μέθοδος επέδειξε βελτιωμένη ακρίβεια τμηματοποίησης, εκπληρώνοντας ποιοτικά κριτήρια και ενισχύοντας την ικανότητα χαρακτηρισμού των ομάδων αποτιτανώσεων με ανάλυση μορφολογίας (μεγέθους και σχήματος) μεμονωμένων αποτιτανώσεων. Οι περιορισμοί της προτεινόμενης μεθόδου τμηματοποίησης αποδίδονται κυρίως: • Στην ανάλυση δύο κλιμάκων του μετασχηματισμού κυματίου, με αποτέλεσμα τον περιορισμό της προσαρμοστικότητας της μεθόδου σε μικροαποτιτανώσεις διαφορετικών μεγεθών. • Στην μέθοδο επέκτασης περιοχής περιοριζόμενη από σημεία περιγράμματος σε 8 διευθύνσεις. Οι περιορισμοί της αξιολόγησης της προτεινόμενης μεθόδου τμηματοποίησης αποδίδονται κυρίως: • Στην ποιοτική μόνο αξιολόγηση της ακρίβειας τμηματοποίησης, μέσω ανάλυσης παρατηρητών. • Στην χρήση περιορισμένου αριθμού χαρακτηριστικών μορφολογίας στο σύστημα αυτόματης διάγνωσης. Για την αντιμετώπιση των προαναφερθέντων περιορισμών, προτάθηκε η μέθοδος Ενεργών Περιγραμμάτων Πολλαπλών Κλιμάκων με αρχικοποίηση Ενεργών Ακτίνων στην αυτόματα επιλεγόμενη αδρή κλίμακα κυματίου. Αρχικά, χρησιμοποιήθηκε ο μετασχηματισμός συνεχούς κυματίου για την παροχή πολλαπλών κλιμάκων ανάλυσης. Στο πεδίο των πολλαπλών κλιμάκων εντοπίζεται η βέλτιστη αδρή κλίμακα (coarse scale) ανάλυσης με βάση τη μέγιστη απόκριση περιοχής μικροαποτιτάνωσης (scale-space MC signature). Στη συγκεκριμένη βέλτιστη κλίμακα απόκρισης εφαρμόζεται η μέθοδος των ενεργών ακτίνων για τον εντοπισμό σημείων του περιγράμματος της μικροαποτιτάνωσης σε 8 διευθύνσεις. Από αυτά τα σημεία ορίζεται πλήρως το περίγραμμα με χρήση μεθόδου γραμμικής παρεμβολής στη βέλτιστη κλίμακα απόκρισης. Κάθε σημείο του περιγράμματος ακολουθεί την κατεύθυνση μεγιστοποίησης της βάθμωσης εικόνας για τον καθορισμό του περιγράμματος στην βέλτιστη κλίμακα (directional Active Contour). Για την τελική εξαγωγή του περιγράμματος, οι θέσεις των σημείων του περιγράμματος επανακαθορίζονται στις κλίμακες μεγαλύτερης ακρίβειας (fine scales). Η ακρίβεια τμηματοποίησης της δεύτερης προτεινόμενης μεθόδου αξιολογήθηκε ποσοτικά με το κριτήριο επικάλυψης περιοχής. Για το σκοπό αυτό χρησιμοποιούνται τμηματοποιήσεις από ειδικευμένο ακτινολόγο. Τμηματοποιήθηκαν συνολικά 1157 μεμονωμένες μικροαποτιτανώσεις προερχόμενες από 128 ομάδες μικροαποτιτανώσεων, ψηφιοποιημένες σε ανάλυση 50μm (βάση δεδομένων DDSM). Μελετήθηκε επίσης η επίδραση της ακρίβειας τμηματοποίησης της δεύτερης προτεινόμενης μεθόδου στην απόδοση μεθόδου αυτόματης διάγνωσης ομάδων αποτιτανώσεων με βάση χαρακτηριστικά ομοιότητας και διαφοροποίησης (variability) ομάδας μικροαποτιτανώσεων, τα οποία περιγράφουν τη μορφολογία μεμονωμένων αποτιτανώσεων (εμβαδού: εμβαδόν, μέγιστη διάμετρος, σχετική αντίθεση, εκκεντρότητα, συμπαγότητα, διακύμανση ακτινικών αποστάσεων, περιοχής: ροπές 1ης και 2ης τάξης, και περιγράμματος: χαρακτηριστικό ροπής και συχνότητας). Ακολούθως, τέσσερα συστήματα αυτόματης διάγνωσης σχεδιάστηκαν βασιζόμενα στον ταξινομητή ελαχίστων τετραγώνων – ελάχιστης απόστασης και μορφολογικά χαρακτηριστικά εξήχθησαν από τις τρεις αυτόματες μεθόδους τμηματοποίησης (δύο προτεινόμενες και μία συγκρινόμενη). Η ποσοτική αξιολόγηση των προτεινόμενων μεθόδων τμηματοποίησης με χρήση δείκτη επικάλυψης περιοχής απέδειξε ότι μόνο η μέθοδος των Ενεργών Περιγραμμάτων Πολλαπλών Κλιμάκων με αρχικοποίηση Ενεργών Ακτίνων στη βέλτιστη κλίμακα ανάλυσης χαρακτηρίζεται από εξίσου υψηλή απόδοση για τις μικρού (<500μm) και μεγάλου (>500μm) μεγέθους μικροαποτιτανώσεις. Επιπλέον, ο ταξινομητής που βασίστηκε σε χαρακτηριστικά εξαγόμενα από τη βελτιστοποιημένη μέθοδο τμηματοποίησης παρουσίασε καλύτερη απόδοση ταξινόμησης (0.779±0.041) από τους ταξινομητές που βασίστηκαν σε χαρακτηριστικά εξαγόμενα από τη μέθοδο Ενεργών Ακτίνων (0.667±0.041) και τη μέθοδο ακτινικής βάθμωσης (0.670±0.044). Η απόδοση ταξινόμησης του βελτιωμένου αλγόριθμου τμηματοποίησης ήταν δε παρόμοια με την απόδοση του ταξινομητή που βασίστηκε σε χαρακτηριστικά εξαγόμενα από χειροκίνητα τμηματοποιημένες μικροαποτιτανώσεις (0.813±0.037). / Accurate segmentation of microcalcifications (MCs) in mammography is crucial for the quantification of morphologic properties by features incorporated in computer-aided diagnosis (CADx) schemes. At first, a novel segmentation method is proposed implementing active rays (polar-transformed active contours) on B-spline wavelet representation to identify microcalcification contour point estimates in a coarse-to-fine strategy at two levels of analysis. An iterative region growing method is used to delineate the final microcalcification contour curve, with pixel aggregation constrained by the microcalcification contour point estimates. A radial gradient method, representing the current state-of-the-art, was also implemented for comparative purposes. The methods were tested on a dataset consisting of 149 mainly pleomorphic microcalcification clusters originating from 130 mammograms of the DDSM database. Segmentation accuracy of both methods was evaluated by three radiologists, based on a 5-point rating scale. The radiologists’ average accuracy ratings were 3.96±0.77, 3.97±0.80 and 3.83±0.89 for the proposed method, and 2.91±0.86, 2.10±0.94 and 2.56±0.76 for the radial gradient-based method, respectively, while the differences in accuracy ratings between the two segmentation methods were statistically significant (Wilcoxon signed-ranks test, p<0.05). The effect of the two segmentation methods in the classification of benign from malignant microcalcification clusters was also investigated. A Least Square Minimum Distance (LSMD) classifier was employed based on cluster features reflecting three morphological properties of individual microcalcifications (area, length and relative contrast). Classification performance was evaluated by means of the area under ROC curve (Az). The area and length morphologic features demonstrated a statistically significant (Mann-Whitney U-test, p<0.05) higher patient-based classification performance when extracted from microcalcifications segmented by the proposed method (0.82±0.06 and 0.86±0.05, respectively), as compared to segmentation by the radial gradient-based method (0.71±0.08 and 0.75±0.08). The proposed method demonstrates improved segmentation accuracy, fulfilling human visual criteria, and enhances the ability of morphologic features to characterize microcalcification clusters. The limitations of this method could be attributed to: • Multiscale analysis restricted to two scales and ad-hoc selection of the coarsest scale of analysis, limiting the desired size-adaptation property of the proposed segmentation method. • Use of constrained region growing to delineate the final MC region to avoid discontinouities inherent to the 8-contour point estimates. • Segmentation accuracy assessed only qualitatively. • Limited morphology anaysis incorporated into the CADx framework. To overcome these limitations, a second method is introduced adaptive to coarse scale selection to initialize the segmentation algorithm, by means of scale-space signatures. Also, we suggest the analysis in the continuous wavelet transform offering a rich multiscale frame. In this framework, multiscale active contours are introduced, utilizing as initial step the previously proposed Active Rays method combined to linear interpolation, for initial contour estimation. Then, each contour point follows the direction where the gradient is maximized. MCs are finally delineated by refining each contour point position at finer scales more accurately. Segmentation accuracy is quantitatively assessed by means of the Area Overlap Measure, utilizing manual segmentation of individual MCs as ground truth, provided by expert radiologists. A total of 1157 individual MCs were segmented in a dataset of 128 MC clusters, digitized at 50μm pixel resolution. To further ensure feature reliability, features extracted from the improved segmentation method were compared (Pearson correlation) to features extracted from manual experts’ delineations. Following, four CADx schemes were implemented utilizing Least Square Minimum Distance (LSMD) classifier and morphology features extracted from the two proposed and the Radial Gradient method. Training of all classifiers was accoblished by features extracted from manually segmented MCs. Quantitative analysis indicated that Multiscale Active Contour method initialized by Active Rays (MAC-AR) had similar Area Overlap Measure performance either for small and large MCs. Furthermore, the improved method demonstrated better performance in terms of classification performance (Az=0.78), as compared to Multiscale Active Rays constrained Region Growing (MAR-RG) (Az=0.67) and the radial gradient one (Az=0.67), however, statistically similar to manual segmentation, representing the best performance (Az=0.81).