As computer technologies are developed rapidly in recent years, the ways to diagnose diseases also alter in clinical practice. Picture Archiving and Communication System (PACS) is an example that makes the diagnostic way for medical images change from view box to monitor. All types of medical images tend to be digitized and this makes it practical for helping doctor diagnose medical images via computer technologies. In this thesis, we propose a systemic approach to screen abnormalities in chest radiographs. First, a preprocess step identifying the view of chest radiographs is introduced. Second, a method is proposed for automated detection of gross abnormal opacity in chest radiographs. Third, computation time reduction is performed by subsampling. Finally, a computer-aided diagnosis system is implemented and evaluated in a clinical environment. Major technique used in this thesis is to analyze the projection profile obtained by projecting a chest image on to the mediolateral axis. The discriminant performance for each method is evaluated by using receiver operating characteristic (ROC) analysis. The results indicate that the proposed methods are potentially useful for screening abnormalities in chest radiographs.

No description available.

Orientador: Sergio Marrone Ribeiro / Resumo: Introdução: Apesar dos avanços na compreensão do acometimento da tuberculose pulmonar na população de transplantados renais, são escassos na literatura os estudos que visam a entender como esta patologia se manifesta através dos métodos de imagem nesta população específica, já que com a imunossupressão podem haver apresentações atípicas de doença, como já é bem estabelecido em outras infecções. Propósito: Estabelecer o número de casos de tuberculose pulmonar ativa na população de transplantados renais de nossa instituição, bem como analisar as manifestações radiológicas desta patologia nas radiografias e nas tomografias computadorizadas de alta resolução de tórax destes pacientes, buscando avaliar padrões de acometimento nestes métodos de imagem e se estes são sobreponíveis ou não à tuberculose pulmonar na população geral Métodos: Foram analisados os prontuários eletrônicos dos pacientes transplantados renais no período de janeiro de 2013 a julho de 2016 em busca de pacientes que tenham apresentado tuberculose pulmonar ativa neste período. Foram colhidos dados do prontuário eletrônico e também analisadas as radiografias e tomografias de tórax nestes pacientes. Resultados: Na população de 769 pacientes transplantados renais de nossa instituição foram encontrados 4 casos de tuberculose pulmonar ativa. As tomografias forneceram informações adicionais às radiografias em 100% dos casos analisados. As manifestações pulmonares da tuberculose pulmonar avaliadas nas tomografias dos qu... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Introduction: Despite the advances in understanding the involvement of pulmonary tuberculosis in the renal transplant population, there are few studies in the literature that aim to understand how this pathology manifests itself through imaging methods in this specific population, since with immunosuppression there may be atypical presentations of disease, as is well established in other infections. Purpose: To establish the number of cases of active tuberculosis in the population of renal transplant patients of our institution, as well as to describe the radiological manifestations of active pulmonary tuberculosis in the thoracic radiography and high resolution computed tomography of patients with active pulmonary tuberculosis, aiming to establish patterns of involvement in these imaging methods and whether they are overlapping or not to the pulmonary tuberculosis in general population. Methods: The electronic medical records of renal transplant patients were analyzed from January 2013 to July 2016 in search of patients who had active pulmonary tuberculosis in this period. Data were collected from the electronic medical record and the thoracic radiography and tomography were also analyzed in these patients. Results: In the population of 769 renal transplant patients from our institution, 4 cases of active pulmonary tuberculosis were found. Tomography provided additional information to radiography in 100% of the cases analyzed. The pulmonary manifestations of pulmonary tuberc... (Complete abstract click electronic access below) / Mestre

Computer Aided Detection (CAD) systems are expecting to gain significant importance in terms of reducing the work load of radiologists and enabling the large screening programs. A large share of CAD systems are based on learning from examples, to enables the decision making between the images with or without disease. Images are simplified to numerical descriptors (features vectors) and the system is trained with these features. The common practical problem with CAD systems is training the system with a data from a specific source and testing it on a data from a different source; the variations between sources usually affect the CAD system function. The possible solutions for this problem are (1) normalizing images to make them look more equal, (2) choosing less variation sensitive features and (3) modifying the classifier so that it classifies the data from different sources more accurately. In this project the effect of image variations on the developed CAD system on chest radio graphs for Tuberculosis is studied at Diagnostic Image Analysis Group. Tuberculosis is one of the major healthcare problems in some parts of the world (1.3 million deaths in 2007) [1]. Although the system has a great performance on the train and test data from the same source, using different sub dataset for training and testing the system does not lead to the same result. To limit the effect of image variation of the CAD systems three different approaches are applied for normalizing the images: (1) Simple normalization, (2) local normalization and (3) multi band local normalization. All three approaches enhance the performance of the system in case of various sub datasets for training and testing purposes. According to the improvement achieved by applying normalization it is suggested as a solution for the stated problem above. Although the outcome of this study has satisfactory result, there is always room for further investigations and studies; in specific testing different approaches for finding less variation sensitive features and modifying the classification procedure to a more variation tolerant process.

Background Machine learning techniques, such as convolutional networks, have shown promise in medical image analysis, including the detection of pediatric pneumonia. However, the interpretability of these models is often lacking, compromising their trustworthiness and acceptance in medical applications. The interpretability of machine learning models in medical applications is crucial for trust and bias identification. Aim The aim is to create a locally interpretable neural network that performs comparably to black-box models while being inherently interpretable, enhancing trust in medical machine learning models. Method An MLP ReLU network is trained with Guangzhou Women and Children's Medical Center pediatric chest x-ray image dataset and utilize Aletheia unwrapper for interpretability. A 5-fold cross-validation assesses the network's performance, measuring accuracy and F1 score. The average accuracy and F1 score are 0.90 and 0.91, respectively. To assessthe interpretability results are compared against a CNN network aided with LIME and SHAP to generate explanations. Results Despite lacking convolutional layers, the MLP network satisfactorily categorizes pneumonia images and explanations align with relevant areas of interest from previous studies. Moreover, by comparing it with a state of the art network aided with LIME and SHAP explanations, the local explanations demonstrate to be consistent within areas of the lungs while the post-hoc alternatives often highlighted areas not relevant for the specific task. Conclusion The developed locally interpretable neural network demonstrates promising performance and interpretability. However, additional research and implementation are required for it to outperform the so-called black box models. In a medical setting, a more accurate model despite the score could be crucial, as it could potentially save more lives, which is the ultimate goal of healthcare.

In recent years, image segmentation by using deep neural networks has made great progress. However, reaching a good result by training with a small amount of data remains to be a challenge. To find a good way to improve the accuracy of segmentation with limited datasets, we implemented a new automatic chest radiographs segmentation experiment based on preliminary works by Chunliang using deep learning neural network combined with shape context information. When the process was conducted, the datasets were put into origin U-net at first. After the preliminary process, the segmented images were then repaired through a new network with shape context information. In this experiment, we created a new network structure by rebuilding the U-net into a 2-input structure and refined the processing pipeline step. In this proposed pipeline, the datasets and shape context were trained together through the new network model by iteration. The proposed method was evaluated on 247 posterior-anterior chest radiographs of public datasets and n-folds cross-validation was also used. The outcome shows that compared to origin U-net, the proposed pipeline reaches higher accuracy when trained with limited datasets. Here the "limited" datasets refer to 1-20 images in the medical image field. A better outcome with higher accuracy can be reached if the second structure is further refined and shape context generator's parameter is fine-tuned in the future. / Under de senaste åren har bildsegmentering med hjälp av djupa neurala nätverk gjort stora framsteg. Att nå ett bra resultat med träning med en liten mängd data kvarstår emellertid som en utmaning. För att hitta ett bra sätt att förbättra noggrannheten i segmenteringen med begränsade datamängder så implementerade vi en ny segmentering för automatiska röntgenbilder av bröstkorgsdiagram baserat på tidigare forskning av Chunliang. Detta tillvägagångssätt använder djupt lärande neurala nätverk kombinerat med "shape context" information. I detta experiment skapade vi en ny nätverkstruktur genom omkonfiguration av U-nätverket till en 2-inputstruktur och förfinade pipeline processeringssteget där bilden och "shape contexten" var tränade tillsammans genom den nya nätverksmodellen genom iteration.Den föreslagna metoden utvärderades på dataset med 247 bröströntgenfotografier, och n-faldig korsvalidering användes för utvärdering. Resultatet visar att den föreslagna pipelinen jämfört med ursprungs U-nätverket når högre noggrannhet när de tränas med begränsade datamängder. De "begränsade" dataseten här hänvisar till 1-20 bilder inom det medicinska fältet. Ett bättre resultat med högre noggrannhet kan nås om den andra strukturen förfinas ytterligare och "shape context-generatorns" parameter finjusteras.