Esse trabalho apresenta uma nova proposta do algoritmo de médias não-locais (NLM - Non-Local Means) para a filtragem do ruído quântico de imagens mamográficas digitais adquiridas com dose de radiação reduzida. A redução nas doses de radiação tem como objetivo principal minimizar os riscos de indução ao câncer de mama causado pela exposição do paciente à radiação ionizante no momento do exame. No entanto, a qualidade da imagem mamográfica diminui com a redução da dose de radiação e o ruído predominante nesse caso é o ruído quântico, que segue a distribuição de Poisson e é dependente do sinal. Como o algoritmo NLM foi originalmente desenvolvido para filtragem de ruído Gaussiano independente do sinal, a proposta desse trabalho foi de adaptar o algoritmo NLM original de modo que ele se tornasse mais adequado para filtragem do ruído encontrado nas imagens mamográficas digitais. Nessa nova abordagem, chamada de Variance Map Non-local Means (VM-NLM), a filtragem do ruído quântico é realizada no próprio domínio da imagem, levando-se em conta a variância do ruído em cada pixel da imagem, já que o ruído é dependente do sinal. Com isso, elimina-se a necessidade de realizar uma estimativa precisa dos parâmetros do ruído para o uso de uma transformada de estabilização de variância (como a transformada generalizada de Anscombe), antes do processo de filtragem. Essa estimativa normalmente requer medidas preliminares no equipamento mamográfico, cujo acesso nem sempre é viável na prática. A proposta foi avaliada em três bancos de imagens mamográficas adquiridas com diferentes doses de radiação. As avaliações de desempenho foram realizadas comparando objetivamente a qualidade das imagens mamográficas obtidas com a dose padrão de radiação com as adquiridas com doses reduzidas, após a filtragem do ruído. Os resultados obtidos com o algoritmo proposto mostraram que ele produz imagens mamográficas mais nítidas e com melhor preservação de bordas e pequenos detalhes do que o algoritmo NLM original. / This work presents a new proposal from the non-local means algorithm (NLM - Non-Local Means) for filtering the quantum noise of digital mammography images acquired with reduced radiation dose. The reduction in radiation doses aims to minimize the risk of inducing breast cancer caused by patient exposure to ionizing radiation during the examination. However, the mammographic image quality decreases with the reduction of the radiation dose and the predominant noise in this case is the quantum noise, which follows the Poisson distribution and it is dependent of the signal. As the NLM algorithm was originally developed for filtering additive Gaussian noise, the purpose of this study was to adapt the original NLM algorithm so that it becomes more suitable for filtering the noise found in digital mammographic images. In this new approach, called Variance Map Non-local Means (VM-NLM), the filtering of the quantum noise is performed in the image domain, considering the noise variance in each pixel of the image, since the noise depends on the pixel value. Thus, it eliminates the need for an accurate estimate of the noise parameters for the use of a variance stabilization transform (such as generalized Anscombe Transformation) before the filtering process. This estimate typically requires preliminary measurements in the mammographic equipment, which is not always viable in clinical practice. The proposal was evaluated in three databases of mammographic images acquired with different radiation doses. Performance evaluations were conducted comparing objectively the quality of mammographic images acquired with standard radiation dose and with reduced doses, after filtering the noise. The results obtained with the proposed algorithm showed that it produces sharper mammographic images with better preservation of edges and small details than the original NLM algorithm.

O objetivo deste trabalho é apresentar um novo método para a remoção do ruído Poisson em imagens de mamografia digital adquiridas com baixa dosagem de radiação. Sabe-se que a mamografia por raios X é o exame mais eficiente para a detecção precoce do câncer de mama, aumentando consideravelmente as chances de cura da doença. No entanto, a radiação absorvida pela paciente durante o exame ainda é um problema a ser tratado. Estudos indicam que a exposição à radiação pode induzir a formação do câncer em algumas mulheres radiografadas. Apesar desse número ser significativamente baixo em relação ao número de mulheres que são salvas pelo exame, existe a necessidade do desenvolvimento de meios que viabilizem a diminuição da dose de radiação empregada. No entanto, uma redução na dose de radiação piora a qualidade da imagem pela diminuição da relação sinal-ruído, prejudicando o diagnóstico médico e a detecção precoce da doença. Nesse sentido, a proposta deste trabalho é apresentar um método para a filtragem do ruído Poisson que é adicionado às das imagens mamográficas quando adquiridas com baixa dosagem de radiação, fazendo com que ela apresente qualidade equivalente àquela adquirida com a dose padrão de radiação. O algoritmo proposto foi desenvolvido baseado em adaptações de algoritmos bem estabelecidos na literatura, como a filtragem no domínio Wavelet, aqui usando o Shrink-thresholding (WTST), e o Block-matching and 3D Filtering (BM3D). Os resultados obtidos com imagens mamográficas adquiridas com phantom e também imagens clínicas, mostraram que o método proposto é capaz de filtrar o ruído adicional incorporado nas imagens sem perda aparente de informação. / The aim of this work is to present a novel method for removing the Poisson noise in digital mammography images acquired with reduced radiation dose. It is known that the X-ray mammography is the most effective exam for early detection of breast cancer, greatly increasing the chances of healing the disease. However, the radiation absorbed by the patient during the exam is still a problem to be treated. Some studies showed that mammography can induce breast cancer in a few women. Although this number is significantly low compared to the number of women who are saved by the exam, it is important to develop methods to enable the reduction of the radiation dose used in the exam. However, dose reduction led to a decrease in image quality by means of the signal to noise ratio, impairing medical diagnosis and the early detection of the disease. In this sense, the purpose of this study is to propose a new method to reduce Poisson noise in mammographic images acquired with low radiation dose, in order to achive the same quality as those acquired with the standard dose. The method is based on well established algorithms in the literature as the filtering in Wavelet domain, here using Shrink-thresholding (WTST) and the Block-matching and 3D Filtering (BM3D). Results using phantom and clinical images showed that the proposed algorithm is capable of filtering the additional noise in images without apparent loss of information.

A proposta deste trabalho foi analisar diferentes metodologias de treinamento de uma rede neural convolucional profunda (CNN) para a detecção de distorção arquitetural mamária (DA) em imagens de mamografia digital. A DA é uma contração sutil do tecido mamário que pode representar o sinal mais precoce de um câncer de mama em formação. Os sistemas computacionais de auxílio ao diagnóstico (CAD) existentes ainda apresentam desempenho insatisfatório para a detecção da DA. Sistemas baseados em CNN têm atraído a atenção da comunidade científica, inclusive na área médica para a otimização dos sistemas CAD. No entanto, as CNNs necessitam de um grande volume de dados para serem treinadas adequadamente, o que é particularmente difícil na área médica. Dessa forma, foi realizada neste trabalho, uma comparação de diferentes abordagens de treinamento para uma arquitetura CNN avaliando-se o efeito de técnicas de geração de novas amostras (data augmentation) sobre o desempenho da rede. Para isso, foram utilizadas 240 mamografias digitais clínicas. Uma das redes (CNN-SW) foi treinada com recortes extraídos por varredura em janela sobre a área interna da mama (aprox. 21600 em média) e a outra rede (CNN-SW+) contou com o mesmo conjunto ampliado por data augmentation (aprox. 345000 em média). Para avaliar o método, foi utilizada validação cruzada por k-fold, gerando-se em rodízio, 10 modelos de cada rede. Os testes analisaram todas as ROIs extraídas da mama, sendo testados 14 mamogramas por fold, e obtendo-se uma diferença estatisticamente significativa entre os resultados (AUC de 0,81 para a CNN-SW e 0,83 para a CNN-SW+). Mapas de calor ilustraram as predições da rede, permitindo uma análise visual e quantitativa do comportamento de ambos os modelos. / The purpose of this work was to analyze different training methodologies of a deep convolutional neural network (CNN) to detect breast architectural distortion (AD) in digital mammography images. AD is a subtle contraction of the breast tissue that may represent the earliest sign of a breast cancer in formation. Current Computer-Aided Detection (CAD) systems still have an unsatisfactory performance on AD detection. CNN-based systems have attracted the attention of the scientific community, including in the medical field for CAD optimization. However, CNNs require a large amount of data to be properly trained, which is particularly difficult in the medical field. Thus, in this work, different training approaches for a CNN architecture are compared evaluating the effect of data augmentation techniques on the data set. For this, 240 clinical digital mammography were used. One of the networks (CNN-SW) was trained with regions of interest (ROI) extracted by a sliding window over the inner breast area (approx 21600 on average) and the other network (CNN-SW+) had the same set enlarged by data augmentation (about 345000 on average). To evaluate the method, k-fold cross-validation was used, generating 10 instances of each model. The tests looked at all the ROIs extracted from the breast (14 mammograms per fold), and results showed a statistically significant difference between both networks (AUC of 0.81 for CNN-SW and 0.83 for CNN-SW+). Heat maps illustrated the predictions of the networks, allowing a visual and quantitative analysis of the behavior of both models.

The aim of this project is the application of Dual Energy technique in breast phantoms using Complimentary-Metal-Oxide-Semiconductor (CMOS)Active-Pixel-Sensor (APS). This includes both, lab experimentation on developed breast phantoms, as well as simulations validating the results. Initially, phantoms were carefully prepared simulating the properties of real breast tissue and were imaged using X-ray unit. The next step in this project involved image processing and data representation. Using the dual energy technique, different concentrations of contrast agent (Iodine) were measured to relate clinical to medium kinetic measurements. With respect to this projects‟ clinical application, the implementation of this technique can be used to evaluate the iodine projected thickness (mg/cm2) using Contrast Enhanced Digital Mammography (CEDM) based on Dual Energy technique for the breast cancer detection. / Ο στόχος του εργασίας αυτής είναι η εφαρμογή της τεχνικής Διπλής Ενέργειας σε ομοιώματα μαστού χρησιμοποιώντας Complimentary-Metal-Oxide-Semiconductor (CMOS) Active-Pixel-Sensor (APS). Αυτό περιλαμβάνει, τον πειραματισμό σε αναπτυγμένα ομοιώματα μαστού, καθώς και προσομοιώσεις για την επικύρωση των αποτελεσμάτων. Αρχικά, κατασκευάστηκαν προσεκτικά τα ομοιώματα μαστού όπου προσομοιώνουν τις πραγματικές ιδιότητες των ιστών του μαστού και στη συνέχεια απεικονίστηκαν με χρήση μονάδων ακτίνων-Χ . Το επόμενο βήμα σε αυτό την εργασία ήταν η επεξεργασία εικόνας και παρουσίαση δεδομένων. Χρησιμοποιώντας την τεχνική της διπλής ενέργειας, διαφορετικές συγκεντρώσεις σκιαγραφικού (ιώδιο) μετρήθηκαν ώστε να σχετίζουν κλινικά την μέτρηση της αγγειογένεσης εν συναρτήση του χρόνου. Η εφαρμογή αυτής της τεχνικής μπορεί να χρησιμοποιηθεί για να αξιολογήσει το προβλεπόμενο πάχος του ιωδίου (mg/cm^2) χρησιμοποιώντας Ενίσχυση Αντίθεσης στην Ψηφιακή Μαστογραφία βασίσμένη στην τεχνική της διπλής ενέργειας για την ανίχνευση του καρκίνου του μαστού.

INTRODUÇÃO: O objetivo deste estudo é determinar a correlação anatomopatológica das calcificações amorfas agrupadas diagnosticadas na mamografia digital de campo total, ou seja, das calcificações suspeitas mais tênues, uma vez que houve aumento da caracterização de calcificações na mamografia digital. MÉTODOS: Estudo retrospectivo baseado nos laudos mamográficos classificados como categoria BI-RADS 4 no primeiro ano de introdução da técnica digital, com análise dos diagnósticos anatomopatológicos das microcalcificações amorfas agrupadas submetidas à biópsia de fragmento assistida à vácuo no serviço. Calculou-se: os achados anatomopatológicos que se associaram ao achado radiológico de microcalcificações amorfas agrupadas e o valor preditivo positivo destas calcificações biopsiadas. RESULTADOS: Dos 219 achados por microcalcificações amorfas agrupadas, 78 foram submetidos à biópsia de fragmento assistida à vácuo com seguimento conhecido ou cirurgia subsequente. O diagnóstico anatomopatológico correspondeu a oito (10%) casos malignos, 36 (46%) casos benignos, e 34 (44%) diagnósticos de lesões de risco, das quais oito (10%) do subgrupo cicatriz radiada/lesões papilíferas (sete cicatrizes radiadas e um papiloma) e 26 (33%) do subgrupo atipias/ neoplasias lobulares, correspondendo a 14 (18%) hiperplasias ductais atípicas, quatro (5%) neoplasias lobulares e oito (10%) lesões de células colunares com atipia (atipia epitelial plana). A ampliação cirúrgica foi recomendada para as lesões com potencial incerto de malignidade à biópsia e realizada em 65% do subgrupo atipias/neoplasias lobulares, com taxa de subestimação nula (0/18). O tempo médio de seguimento das pacientes com diagnóstico benigno ou de lesão de risco foi 22 meses. CONCLUSÕES: Um terço das microcalcificações amorfas agrupadas em mamografia digital de campo total corresponderam a lesões precursoras representadas pelas atipias (ductais e colunares) e neoplasias lobulares. Essas lesões se associaram às calcificações amorfas agrupadas numa proporção de cerca de 3:1 em relação às lesões malignas, cujo VPP 3 foi 10% / PURPOSE: To determine the anatomopathological correlation of grouped amorphous calcifications (the most tenuous of suspicious calcifications) disclosed on full-field digital mammography, given the enhanced characterization of calcifications provided by digital mammography. METHODS: A retrospective study of mammographic reports classified as BI-RADS® category 4 at a private diagnostic service specialized in breast imaging was carried out on exams performed during the first year of introducing the digital technique. The investigation entailed analysis of the anatomopathological diagnoses of BI-RADS® category 4 for grouped amorphous microcalcifications submitted to vacuum-assisted breast biopsy (VABB). Anatomopathological findings correlated to this radiological finding were determined and positive predictive value of these calcifications submitted to biopsy (PPV 3) was calculated. RESULTS: Of the 219 findings of grouped amorphous microcalcifications, 78 were submitted to VABB with known follow-up or subsequent surgery. The anatomopathological results included eight (10%) malignant cases, 36 (46%) benign cases and 34 (44%) diagnoses of high-risk lesions, eight of which belonged to the radial scar/papillary lesion subgroup (seven radial scars and one papilloma) and 26 (33% of all cases) to the atypia/lobular neoplasia subgroup, comprising 14 atypical ductal hyperplasias, four lobular neoplasias and eight flat epithelial atypia. Extended surgery was recommended for lesions with uncertain malignant potential at biopsy and performed in 65% of the atypia/lobular neoplasia subgroup, with an underestimation rate of zero (0/18). Mean follow-up time of patients diagnosed with benign or high-risk lesions was 22 months. CONCLUSIONS: One-third of grouped amorphous calcifications on full-field digital mammography corresponded to precursory lesions in the form of atypia (ductal and columnar) or lobular neoplasias. These lesions were associated to grouped amorphous calcifications at a ratio of 3:1 compared to malignant lesions,whose PPV 3 was 10%

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-07-01T12:22:19Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Filipe_Cordeiro.pdf: 19608976 bytes, checksum: a0ff2fa1256af4323f10bfcbb3df974d (MD5) / Made available in DSpace on 2016-07-01T12:22:19Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Filipe_Cordeiro.pdf: 19608976 bytes, checksum: a0ff2fa1256af4323f10bfcbb3df974d (MD5) Previous issue date: 2015-12-16 / De acordo com a Organização Mundial de Saúde, o câncer de mama é a forma mais comum de câncer entre as mulheres no mundo todo, sendo um dos tipos de câncer mais fatal. Estudos mostram que o diagnóstico precoce pode contribuir para a redução da taxa de mortalidade e aumentar as opções de tratamento. Apesar da existência de várias técnicas de obtenção de imagens no auxílio ao diagnóstico de câncer de mama, a mamografia digital é ainda a tecnologia mais eficaz e utilizada para esse fim. Consequentemente, a segmentação de imagens de mamografia é uma tarefa fundamental para auxiliar o diagnóstico, levando em consideração a forma da lesão mamária e suas bordas. No entanto, a segmentação de imagens de mamografia é uma tarefa complexa, uma vez que ela é muito dependente dos tipos de tecido mamário e da lesão. O algoritmo GrowCut é um método de segmentação de propósito geral baseado em autômatos celulares, capaz de realizar uma segmentação precisa através da seleção adequada de pontos internos e externos à região de interesse. Neste trabalho é apresentado um novo algoritmo semi-supervisionado baseado na modificação do algoritmo GrowCut para realizar segmentação de imagens de mamografia de forma semi-automática. No método proposto é utilizada uma função de pertinência fuzzy Gaussiana para modificar a regra de evolução do algoritmo GrowCut original, visando estimar as probabilidades de um pixel pertencer ao objeto ou fundo da imagem. Esse modelo permite uma maior flexibilidade na inicialização das sementes quando comparado à trabalhos no estado da arte, pois a marcação realizada pelo especialista é utilizada extraindo-se informação do conjunto de sementes, e não informações do posicionamento individual, como o presente no GrowCut clássico. Foi também desenvolvido uma etapa de geração automática de sementes, onde apenas pontos internos da região de interesse são selecionados, através do uso do método de otimização Evolução Diferencial. Além disso, foi desenvolvido um método de ajuste de parâmetros adaptativo, que a partir da extração de características da imagem ajusta os melhores parâmetros para o algoritmo. A abordagem desenvolvida é comparada qualitativamente e quantitativamente com técnicas de segmentação do estado da arte BEMD, BMCS, WAGA, Abordagem Topográfica e MCW, usando métricas relacionadas à forma das regiões segmentadas. As análises são avaliadas utilizando regiões de interesse da base IRMA, totalizando 1.165 mamogramas. Resultados mostram que o algoritmo proposto obteve melhores resultados, considerando similaridade com imagem ouro, para as métricas utilizadas. Para validar a proposta , foi construído um classificador de imagem usando o Perceptron Multicamadas clássico. Resultados mostraram que a técnica proposta obteve taxa de classificação de 94,77%, evidenciado a viabilidade do método proposto. / According to the World Health Organization, breast cancer is the most common cancer in women worldwide, becoming one of the most fatal types of cancer. Several studies show that the early diagnosis technologies can contribute to reduce the mortality rates and improve treatment options. Despite the existence of several imaging techniques to aid at the diagnosis of breast cancer, digital mammography is still the most used and effective imaging technology. Consequently, mammographic image segmentation is a fundamental task to support image diagnosis, considering shape analysis of mammary lesions and their borders. However, mammogram image segmentation is a hard task, once it is highly dependent on the types of mammary tissues. The GrowCut algorithm is a general-purpose segmentation method based on cellular automata, able to perform accurate segmentation through the adequate selection of internal and external points of the region of interest. Herein this work we present a new semi-supervised segmentation algorithm based on the modification of the GrowCut algorithm to perform semi-automatic mammographic image segmentation. In our proposal, we used a fuzzy Gaussian membership function to modify the evolution rule of the original GrowCut algorithm, in order to estimate the probabilities of a pixel being object or background. This model allows flexibility in the seeds initialization when compared to state of the art techniques, because the annotation executed by the specialist is used through the extraction of information of set os seeds, in opposite to the individual seeds information present in classical GrowCut .An automatic seed generation step was developed, where only the seeds internal to the region of interest are selected, using the Differential Evolution algorithm. Furthermore, we developed an adaptive parameter tuning method, which from the image characteristics it find the best parameters to the algorithm. The proposed approach was qualitatively and quantitatively compared with other state-of-the-art segmentation techniques BEMD, BMCS, WAGA, Topographic Approach and MCW, using metrics related to the shape of segmented regions. The analysis are evaluated using regions of interest from IRMA database, totaling 1.165 mammograms. Results show that the proposed algorithm achieved better results, considering similarity with ground truth, for the used metrics. In order to validate our proposal we built an image classifier using a classical Multilayer Perceptron. This analysis employed 1.165 mammograms from IRMA breast cancer database Results show that the proposed technique could achieve a classification rate of 94.77%, evidencing the feasibility of our approach.

In the last 10-15 years there has been a renewed interest in amorphous Se (a-Se) and its alloys due to their application as photoconductor materials in the new fully digital direct conversion flat panel x-ray medical image detectors. For a number of reasons, the a-Se photoconductor layer in such x-ray detectors has to be operated at very high electric fields (up to 10 Volts per micron) and one of the most difficult problems related to such applications of a Se is the problem of the dark current (the current in the absence of any radiation) minimization in the photoconductor layer. <p>This PhD work has been devoted to researching the possibilities for dark current minimization in a-Se x-ray photoconductors devices through a systematic study of the charge transport (carrier mobility and carrier lifetimes) and dark currents in single and multilayered a-Se devices as a function of alloying, doping, deposition condition and other fabrication factors. The results of the studies are extensively discussed in the thesis. We have proposed a new technological method for dark current reduction in single and multilayered a-Se based photoconductor for x-ray detector applications. The new technology is based on original experimental findings which demonstrate that both hole transport and the dark currents in a-Se films are a very strong function of the substrate temperature (Tsubstrate) during the film deposition process. We have shown that the new technique reduces the dark currents to approximately the same levels as achievable with the previously existing methods for dark current reduction. However, the new method is simpler to implement, and offers some potential advantages, especially in cases when a very high image resolution (20 cycles/mm) and/or fast pixel readout (more than 30 times per second) are needed. <p>Using the new technology we have fabricated simple single and double (ni-like) photoconductor layers on prototype x-ray image detectors with CCD (Charge Coupled Device) readout circuits. Dark currents in the a-Se photoconductor layer were not a problem for detector operation at all tested electric fields. Compared to the currently available commercial systems for mammography, the prototype detectors have demonstrated an excellent imaging performance, in particular superior spatial resolution (20 cycles/mm). Thus, the newly proposed technology for dark current reduction has shown a potential for commercialization.

In the last 10-15 years there has been a renewed interest in amorphous Se (a-Se) and its alloys due to their application as photoconductor materials in the new fully digital direct conversion flat panel x-ray medical image detectors. For a number of reasons, the a-Se photoconductor layer in such x-ray detectors has to be operated at very high electric fields (up to 10 Volts per micron) and one of the most difficult problems related to such applications of a Se is the problem of the dark current (the current in the absence of any radiation) minimization in the photoconductor layer. <p>This PhD work has been devoted to researching the possibilities for dark current minimization in a-Se x-ray photoconductors devices through a systematic study of the charge transport (carrier mobility and carrier lifetimes) and dark currents in single and multilayered a-Se devices as a function of alloying, doping, deposition condition and other fabrication factors. The results of the studies are extensively discussed in the thesis. We have proposed a new technological method for dark current reduction in single and multilayered a-Se based photoconductor for x-ray detector applications. The new technology is based on original experimental findings which demonstrate that both hole transport and the dark currents in a-Se films are a very strong function of the substrate temperature (Tsubstrate) during the film deposition process. We have shown that the new technique reduces the dark currents to approximately the same levels as achievable with the previously existing methods for dark current reduction. However, the new method is simpler to implement, and offers some potential advantages, especially in cases when a very high image resolution (20 cycles/mm) and/or fast pixel readout (more than 30 times per second) are needed. <p>Using the new technology we have fabricated simple single and double (ni-like) photoconductor layers on prototype x-ray image detectors with CCD (Charge Coupled Device) readout circuits. Dark currents in the a-Se photoconductor layer were not a problem for detector operation at all tested electric fields. Compared to the currently available commercial systems for mammography, the prototype detectors have demonstrated an excellent imaging performance, in particular superior spatial resolution (20 cycles/mm). Thus, the newly proposed technology for dark current reduction has shown a potential for commercialization.