A segmentação é uma etapa intermediária no registro e reconstrução 3D do pulmão. Geralmente, os métodos de segmentação são interativos e utilizam diferentes estratégias para combinar a expertise dos humanos e a velocidade e precisão dos computadores. A segmentação de imagens RM do pulmão é particularmente difícil devido à grande variação na qualidade da imagem. Dois métodos para a segmentação do contorno do pulmão são apresentados. No primeiro, uma análise individual de cada imagem da série de imagens RM é realizada, e a segmentação ocorre através de técnicas de limiarização e labeling. No segundo método, a respiração é associada a uma função respiração padrão, e através de técnicas de processamento de imagem 2D, detecção de bordas e transformada de Hough, padrões respiratórios são obtidos e, conseqüentemente, a posição dos pontos no tempo são estimados. Seqüências temporais de imagens RM são segmentadas, considerando a coerência no tempo. Desta forma, a silhueta do pulmão pode ser determinada em cada quadro, mesmo em quadros com bordas obscuras. A região do pulmão é segmentada em três etapas, neste método: uma máscara contendo a região do pulmão é criada a partir do resultado do primeiro método de segmentação; a transformada de Hough é aplicada exclusivamente aos pixels da máscara em diversos planos; o contorno do pulmão é extraído do resultado da transformada de Hough utilizando os contornos ativos. O formato da máscara pode ter uma grande variação, e a transformada de Hough modificada pode lidar com essa variação. Os resultados obtidos pelos dois métodos são comparados. / The segmentation of the lung is an intermediary step towards its registry and 3D reconstruction. Usually, segmentation methods are interactive and make use of different strategies to combine the expertise of the human and the computers accuracy and speed. Segmentation of lung magnetic resonance (MR) images is particularly difficult because of the large variation in image quality. Two methods for the lung contour segmentation are presented. In this first method, an individual analysis of each image in the series approach is taken, and the segmentation is made through thresholding and labeling techniques. In the second method, the breathing is associated to a standard respiratory function, and through 2D image processing, edge detection and Hough transform, respiratory patterns are obtained and, consequently, the position of points in time are estimated. Temporal sequences of MR images are segmented by considering the coherence in time. This way, the lung silhouette can be determined in every frame, even on frames with obscure edges. The lung region is segmented in three steps: a mask containing the lung region is created from the results of the first method; the Hough transform is applied exclusively to mask pixels in different planes; and the lung contour is created from the results of the Hough transform through active contours. The shape of the mask can have a large variation, and the modified Hough transform can handle such a shape variation. Results from both methods are compared.

A segmentação é uma etapa intermediária no registro e reconstrução 3D do pulmão. Geralmente, os métodos de segmentação são interativos e utilizam diferentes estratégias para combinar a expertise dos humanos e a velocidade e precisão dos computadores. A segmentação de imagens RM do pulmão é particularmente difícil devido à grande variação na qualidade da imagem. Dois métodos para a segmentação do contorno do pulmão são apresentados. No primeiro, uma análise individual de cada imagem da série de imagens RM é realizada, e a segmentação ocorre através de técnicas de limiarização e labeling. No segundo método, a respiração é associada a uma função respiração padrão, e através de técnicas de processamento de imagem 2D, detecção de bordas e transformada de Hough, padrões respiratórios são obtidos e, conseqüentemente, a posição dos pontos no tempo são estimados. Seqüências temporais de imagens RM são segmentadas, considerando a coerência no tempo. Desta forma, a silhueta do pulmão pode ser determinada em cada quadro, mesmo em quadros com bordas obscuras. A região do pulmão é segmentada em três etapas, neste método: uma máscara contendo a região do pulmão é criada a partir do resultado do primeiro método de segmentação; a transformada de Hough é aplicada exclusivamente aos pixels da máscara em diversos planos; o contorno do pulmão é extraído do resultado da transformada de Hough utilizando os contornos ativos. O formato da máscara pode ter uma grande variação, e a transformada de Hough modificada pode lidar com essa variação. Os resultados obtidos pelos dois métodos são comparados. / The segmentation of the lung is an intermediary step towards its registry and 3D reconstruction. Usually, segmentation methods are interactive and make use of different strategies to combine the expertise of the human and the computers accuracy and speed. Segmentation of lung magnetic resonance (MR) images is particularly difficult because of the large variation in image quality. Two methods for the lung contour segmentation are presented. In this first method, an individual analysis of each image in the series approach is taken, and the segmentation is made through thresholding and labeling techniques. In the second method, the breathing is associated to a standard respiratory function, and through 2D image processing, edge detection and Hough transform, respiratory patterns are obtained and, consequently, the position of points in time are estimated. Temporal sequences of MR images are segmented by considering the coherence in time. This way, the lung silhouette can be determined in every frame, even on frames with obscure edges. The lung region is segmented in three steps: a mask containing the lung region is created from the results of the first method; the Hough transform is applied exclusively to mask pixels in different planes; and the lung contour is created from the results of the Hough transform through active contours. The shape of the mask can have a large variation, and the modified Hough transform can handle such a shape variation. Results from both methods are compared.

No description available.

O estudo do movimento pulmonar é assunto de grande interesse na área médica. A observação direta do mesmo é inviável, uma vez que o pulmão colapsa quando a caixa torácica é aberta. Dentre os meios de observação indireta, escolheu-se o imageamento por ressonância magnética em respiração livre e sem uso de nenhum gás para melhorar o contraste ou qualquer informação de sincronismo. Esta escolha propõe diversos desafios, como: a superar a alta variação na qualidade das imagens, que é baixa, em geral, e a suscetibilidade a artefatos, entre outras limitações a serem superadas. Imagens de Tomografia Computadorizada apresentam melhor qualidade e menor tempo de aquisição, mas expõem o paciente a níveis consideráveis de radiação ionizante. É apresentada uma metodologia para segmentação do pulmão, produzindo um conjunto de pontos coordenados. Isto é feito através do processamento temporal da sequência de imagens de RM. Este processamento consiste nas seguintes etapas: geração de imagens temporais (2DSTI), transformada de Hough modificada, algoritmo de contornos ativos e geração de silhueta. A partir de um dado ponto, denominado centro de rotação, são geradas diversas imagens temporais com orientações variadas. É proposta uma formulação modificada da transformada de Hough para determinar curvas parametrizadas que sejam síncronas ao movimento diafragmático, chamados movimentos respiratórios. Também são utilizadas máscaras para delimitar o domínio de aplicação da transformada de Hough. São obtidos movimentos respiratórios que são suavizados pelo algoritmo de contornos ativos e, assim, permitem a geração de contornos para cada quadro pertencente a sequência e, portanto, de uma silhueta do pulmão para cada sequência. / Lung movement visualization is of great interest in medicine. Direct observation of the lung movement is not practicable, as it collapses if the thoracic cage is opened. Among indirect observation means, we choose magnetic resonance imaging, acquired on free breathing, without the use of any triggering information and any special gas to enhance the contrast. This choice leads us to overcome the high variation on MR images\' quality, which is, generally, low, and, also, artifact susceptibility, among other limitations. Computed Tomography images have better quality and a shorter acquisition time, but they expose the subject to considerably high levels of radiation. A Lung segmentation methodology is presented and it produces a connected set of points. That is achieved through MRI sequences temporal processing and it consists of the following stages: masks generation, 2-dimensional space-time images (2DSTI), modified Hough transform, an active contours algorithm and silhouette generation. Using a given point, which will be called parameter point, various temporal images with varied orientation will be generated. A modified Hough transform is applied to extract parameterized curves, that are synchronous to diaphragmatic movement, which will be called respiratory movements. Also, masks will be used in order to delimit the modified Hough transform application domain. An active contours algorithm will smoothen the obtained respiratory movements, so they will allow the generation of a contour for each frame on the image sequence and, therefore, a lung silhouette for a given sequence.

Optimal surface finding (OSF), a graph-based optimization approach to image segmentation, represents a powerful framework for medical image segmentation and analysis. In many applications, a pre-segmentation is required to enable OSF graph construction. Also, the cost function design is critical for the success of OSF. In this thesis, two issues in the context of OSF segmentation are addressed. First, a robust model-based segmentation method suitable for OSF initialization is introduced. Second, an OSF-based segmentation refinement approach is presented. For segmenting complex anatomical structures (e.g., lungs), a rough initial segmentation is required to apply an OSF-based approach. For this purpose, a novel robust active shape model (RASM) is presented. The RASM matching in combination with OSF is investigated in the context of segmenting lungs with large lung cancer masses in 3D CT scans. The robustness and effectiveness of this approach is demonstrated on 30 lung scans containing 20 normal lungs and 40 diseased lungs where conventional segmentation methods frequently fail to deliver usable results. The developed RASM approach is generally applicable and suitable for large organs/structures. While providing high levels of performance in most cases, OSF-based approaches may fail in a local region in the presence of pathology or other local challenges. A new (generic) interactive refinement approach for correcting local segmentation errors based on the OSF segmentation framework is proposed. Following the automated segmentation, the user can inspect the result and correct local or regional segmentation inaccuracies by (iteratively) providing clues regarding the location of the correct surface. This expert information is utilized to modify the previously calculated cost function, locally re-optimizing the underlying modified graph without a need to start the new optimization from scratch. For refinement, a hybrid desktop/virtual reality user interface based on stereoscopic visualization technology and advanced interaction techniques is utilized for efficient interaction with the segmentations (surfaces). The proposed generic interactive refinement method is adapted to three applications. First, two refinement tools for 3D lung segmentation are proposed, and the performance is assessed on 30 test cases from 18 CT lung scans. Second, in a feasibility study, the approach is expanded to 4D OSF-based lung segmentation refinement and an assessment of performance is provided. Finally, a dual-surface OSF-based intravascular ultrasound (IVUS) image segmentation framework is introduced, application specific segmentation refinement methods are developed, and an evaluation on 41 test cases is presented. As demonstrated by experiments, OSF-based segmentation refinement is a promising approach to address challenges in medical image segmentation.

O estudo do movimento pulmonar é assunto de grande interesse na área médica. A observação direta do mesmo é inviável, uma vez que o pulmão colapsa quando a caixa torácica é aberta. Dentre os meios de observação indireta, escolheu-se o imageamento por ressonância magnética em respiração livre e sem uso de nenhum gás para melhorar o contraste ou qualquer informação de sincronismo. Esta escolha propõe diversos desafios, como: a superar a alta variação na qualidade das imagens, que é baixa, em geral, e a suscetibilidade a artefatos, entre outras limitações a serem superadas. Imagens de Tomografia Computadorizada apresentam melhor qualidade e menor tempo de aquisição, mas expõem o paciente a níveis consideráveis de radiação ionizante. É apresentada uma metodologia para segmentação do pulmão, produzindo um conjunto de pontos coordenados. Isto é feito através do processamento temporal da sequência de imagens de RM. Este processamento consiste nas seguintes etapas: geração de imagens temporais (2DSTI), transformada de Hough modificada, algoritmo de contornos ativos e geração de silhueta. A partir de um dado ponto, denominado centro de rotação, são geradas diversas imagens temporais com orientações variadas. É proposta uma formulação modificada da transformada de Hough para determinar curvas parametrizadas que sejam síncronas ao movimento diafragmático, chamados movimentos respiratórios. Também são utilizadas máscaras para delimitar o domínio de aplicação da transformada de Hough. São obtidos movimentos respiratórios que são suavizados pelo algoritmo de contornos ativos e, assim, permitem a geração de contornos para cada quadro pertencente a sequência e, portanto, de uma silhueta do pulmão para cada sequência. / Lung movement visualization is of great interest in medicine. Direct observation of the lung movement is not practicable, as it collapses if the thoracic cage is opened. Among indirect observation means, we choose magnetic resonance imaging, acquired on free breathing, without the use of any triggering information and any special gas to enhance the contrast. This choice leads us to overcome the high variation on MR images\' quality, which is, generally, low, and, also, artifact susceptibility, among other limitations. Computed Tomography images have better quality and a shorter acquisition time, but they expose the subject to considerably high levels of radiation. A Lung segmentation methodology is presented and it produces a connected set of points. That is achieved through MRI sequences temporal processing and it consists of the following stages: masks generation, 2-dimensional space-time images (2DSTI), modified Hough transform, an active contours algorithm and silhouette generation. Using a given point, which will be called parameter point, various temporal images with varied orientation will be generated. A modified Hough transform is applied to extract parameterized curves, that are synchronous to diaphragmatic movement, which will be called respiratory movements. Also, masks will be used in order to delimit the modified Hough transform application domain. An active contours algorithm will smoothen the obtained respiratory movements, so they will allow the generation of a contour for each frame on the image sequence and, therefore, a lung silhouette for a given sequence.

No description available.

This thesis deals with the segmentation of lung parenchyma and extraction of airways tree from three dimensional CT scans. The external mask of the lungs is created and subsequently used to ease the process of airway segmentation. In this work, some published methods for airways segmentation are described with focus on one, which is described more in detail and also implemented in MATLAB. The proposed approach is based on morphological grayscale reconstruction. Method is tested on the real patient CT scans and finally, the results are discussed.

Svenska socialstyrelsen presenterade år 2017 att lungcancer är den vanligaste cancerrelaterade dödsorsaken bland kvinnor i Sverige och den näst vanligaste bland män. Ett sätt att ta reda på om en patient har lungcancer är att en läkare studerar en tredimensionell-röntgenbild av en patients lungor. För att förebygga misstag som kan orsakas av den mänskliga faktorn är det möjligt att använda datorer och avancerade algoritmer för att upptäcka lungcancer. En nätverksmodell kan tränas att upptäcka detaljer och avvikelser i en lungröntgenbild, denna teknik kallas deep structural learning. Det är både tidskrävande och avancerat att skapa en sådan modell, det är därför viktigt att modellen tränas korrekt. Det finns flera studier som behandlar olika nätverksarkitekturer, däremot inte vad förbehandlingstekniken lung-segmentering kan ha för inverkan på en modell av denna signifikans. Därför ställde vi frågan: hur påverkas accuracy och loss hos en konvolutionell nätverksmodell när lung-segmentering appliceras på modellens tränings- och testdata? För att besvara frågan skapade vi flera modeller som använt, respektive, inte använt lung-segmentering. Modellernas resultat evaluerades och jämfördes, tekniken visade sig motverka överträning. Vi anser att denna studie kan underlätta för framtida forskning inom samma och liknande problemområde. / In the year of 2017 the Swedish social office reported the most common cancer related death amongst women was lung cancer and the second most common amongst men. A way to find out if a patient has lung cancer is for a doctor to study a computed tomography scan of a patients lungs. This introduces the chance for human error and could lead to fatal consequences. To prevent mistakes from happening it is possible to use computers and advanced algorithms for training a network model to detect details and deviations in the scans. This technique is called deep structural learning. It is both time consuming and highly challenging to create such a model. This discloses the importance of decorous training, and a lot of studies cover this subject. What these studies fail to emphasize is the significance of the preprocessing technique called lung segmentation. Therefore we investigated how is the accuracy and loss of a convolutional network model affected when lung segmentation is applied to the model’s training and test data? In this study a number of models were trained and evaluated on data where lung segmentation was applied, in relation to when it was not. The final conclusion of this report shows that the technique counteracts overfitting of a model and we allege that this study can ease further research within the same area of study.

The initial stage of a lung examination involves the segmentation of a CT image, a process that has been put under a lot of pressure with the high demand for chest scans and accurate segmentations. Current automatic segmentation algorithms are either non-robust for different datasets, not easily accessible, or time-consuming. Furthermore, classification of lung diseases such as IPF and NSIP is a difficult task often requiring decision-making between pathologists, radiologists and clinicians to make an accurate prognosis.  Therefore, this thesis aims to create two algorithms easily accessible through a common medical software, 3D Slicer, with simple user interfaces for more efficient lung analysis. The first one is a fully automatic segmentation algorithm with a manual adjustment option. It is robust and developed on a diverse dataset, demonstrating a high accuracy with a median Dice score of0,967. The second one is a lung vessel tree morphometry algorithm which computes various parameters correlated to the vessel tree and its structure, providing insight into morphological changes. It shows great usability but has certain limitations, making it not entirely finished for clinical research but acts as an excellent starting point for a future project. The segmentation algorithm was developed using classical image processing techniques making it comprehensible. The distinctive feature of this algorithm is the entropy map used, enabling an effective way in distinguishing between the fibrotic regions of the lungs with surrounding soft tissue and therefore increasing its applicability on lungs with various diseases. The lung vessel tree morphometry algorithm utilized a segmentation of the lung vessels to organize them into a tree-like structure. The structure was divided into branches where each branch was used to calculate different parameters such as its level within the tree hierarchy, the length of the branch and more. These parameters were displayed and color-coded for further analysis. The obtained result underscores the substantial potential and importance of these developed algorithms for clinical research by providing user-friendly, robust and reliable methods. / Det inledande skedet av en lungundersökning involverar segmenteringen av en CT-bild, en process som har satts under mycket press på grund utav den höga efterfrågan på bröstskanningar och noggrann segmentering. Aktuella automatiska segmenteringsalgoritmer är antingen icke-robusta för olika dataset, ej lättillgängliga eller tidskrävande. Dessutom är klassificering av lungsjukdomar som IPF och NSIP en svår uppgift som ofta kräver beslutsfattande mellan patologer, radiologer och kliniker för att göra en korrekt prognos. Därför syftar denna rapport till att skapa två lättillgängliga algoritmer genom en ofta användmedicinsk programvara, 3D Slicer, bestående utav enkla användargränssnitt för en effektivare analys av lungorna. Den första är en helautomatisk segmenteringsalgoritm med ett manuellt justeringsalternativ. Den är robust och utvecklad på ett mångsidigt dataset som har demonstrerat en hög noggrannhet med en median Dice-score på 0,967. Den andra är en morfometri algoritm för lungkärlsträd som beräknar olika parametrar korrelerade till kärlträdet och dess struktur, vilket ger insikt i morfologiska förändringar. Den visar stor användbarhet men innehåller begränsningar, vilket gör den ej helt färdig för klinisk forskning utan fungerar som en utmärkt utgångspunkt för framtida arbete. Segmenteringsalgoritmen utvecklades med hjälp av klassiska bildbehandlingsmetoder vilket gör den mer lättförstådd. Det utmärkande för denna algoritm är entropikartan som används, vilket möjliggör ett effektivt sätt att skilja mellan de fibrotiska regionerna i lungorna med omgivande mjukdelar, detta gör den mer användbar på lungor med olika sjukdomar. Algoritmen för lungkärlsträdets morfometri använde en segmentering av lungkärlen för att sedanorganiseras i en trädliknande struktur. Strukturen var uppdelad i grenar där varje gren användes för att beräkna olika parametrar såsom dess nivå inom trädhierarkin, grenens längd med mera. Dessutom uppvisades dessa parametrar och färgkodades för vidare analys. Det erhållna resultatet understryker den substantiella potential och betydelse som dessa utvecklade algoritmer kommer att ha i klinisk forskning genom att tillhandahålla användarvänliga, robusta och pålitliga metoder