Improving the generalization of deep learning methods to segment the left ventricle in short axis MR images.

Cardiovascular disease is one of the major health problems worldwide. In clinical practice, cardiac magnetic resonance imaging (CMR) is considered the gold-standard imaging modality for the evaluation of the function and structure of the left ventricle (LV). More recently, deep learning methods have been used to segment LV with impressive results. On the other hand, this kind of approach is prone to overfit the training data, and it does not generalize well between different data acquisition centers, thus creating constraints to the use in daily routines. In this paper, we explore methods to improve the generalization in the segmentation performed by a convolutional neural network. We applied a U-net based architecture and compared two different pre-processing methods to improve uniformity in the image contrast between five cross-dataset training and testing. Overall, we were able to perform the segmentation of the left ventricle using multiple cross-dataset combinations of train and test, with a mean endocardium dice score of 0.82.Clinical Relevance- This work improves the result between the cross-dataset evaluation of the left ventricle segmentation, reducing the constraints for daily clinical adoption of a fully-automatic segmentation method.

Fully Automated Quantification of Cardiac Indices from Cine MRI Using a Combination of Convolution Neural Networks.

Cardiovascular magnetic resonance imaging (CMRI) is one of the most accurate non-invasive modalities for evaluation of cardiac function, especially the left ventricle (LV). In this modality, the manual or semi-automatic delineation of LV by experts is currently the standard clinical practice for chambers segmentation. Despite these efforts, global quantification of LV remains a challenge. In this work, a combination of two convolutional neural network (CNN) architectures for quantitative evaluation of the LV is described, which estimates the cavity and the myocardium areas, endocardial cavity dimensions in three directions, and the myocardium regional wall thickness in six radial directions. The method was validated in CMRI exams of 56 patients (LVQuan19 dataset) and evaluated by metrics Dice Index, Mean Absolute Error, and Correlation with superior performance compared to the state-of-the-art methods. The combination of the CNN architectures provided a simpler yet fully automated approach, requiring no specialist interaction.Clinical Relevance- With the proposed method, it is possible to perform automatically the full quantification of regional clinically relevant parameters of the left ventricle in short-axis CMRI images with superior performance compared to state-of-the-art methods.

Georeferenced and secure mobile health system for large scale data collection in primary care.

INTRODUCTION: Mobile health consists in applying mobile devices and communication capabilities for expanding the coverage and improving the effectiveness of health care programs. The technology is particularly promising for developing countries, in which health authorities can take advantage of the flourishing mobile market to provide adequate health care to underprivileged communities, especially primary care. In Brazil, the Primary Care Information System (SIAB) receives primary health care data from all regions of the country, creating a rich database for health-related action planning. Family Health Teams (FHTs) collect this data in periodic visits to families enrolled in governmental programs, following an acquisition procedure that involves filling in paper forms. This procedure compromises the quality of the data provided to health care authorities and slows down the decision-making process. OBJECTIVES: To develop a mobile system (GeoHealth) that should address and overcome the aforementioned problems and deploy the proposed solution in a wide underprivileged metropolitan area of a major city in Brazil. METHODS: The proposed solution comprises three main components: (a) an Application Server, with a database containing family health conditions; and two clients, (b) a Web Browser running visualization tools for management tasks, and (c) a data-gathering device (smartphone) to register and to georeference the family health data. A data security framework was designed to ensure the security of data, which was stored locally and transmitted over public networks. RESULTS: The system was successfully deployed at six primary care units in the city of Sao Paulo, where a total of 28,324 families/96,061 inhabitants are regularly followed up by government health policies. The health conditions observed from the population covered were: diabetes in 3.40%, hypertension (age >40) in 23.87% and tuberculosis in 0.06%. This estimated prevalence has enabled FHTs to set clinical appointments proactively, with the aim of confirming or detecting cases of non-communicable diseases more efficiently, based on real-time information. CONCLUSION: The proposed system has the potential to improve the efficiency of primary care data collection and analysis. In terms of direct costs, it can be considered a low-cost solution, with an estimated additional monthly cost of U$ 0.040 per inhabitant of the region covered, or approximately U$ 0.106 per person, considering only those currently enrolled in the system.

Ambiente para extração de informações de saúde a partir de bases de dados do SUS / Environment for extracting health information starting from the SUS database

A aplicação de técnicas para produção de informação gerencial e descoberta de conhecimentos em grandes bases de dados, como as existentes nos sistemas de informação do DATASUS, pode representar um avanço substancial na gestão do Sistema Único de Saúde (SUS). Os dados da saúde pública são produzidos por vários sistemas isolados e não integrados, tornando mais difícil a tarefa de produzir informação gerencial. Essa dificuldade motivou este trabalho, cujo objetivo é criar um ambiente (MinerSUS) para extração de informação, a partir da mineração das bases de dados do SUS no Estado de São Paulo. Nesse sentido, foi implantado um ambiente adequado às peculiaridades da Saúde Pública e dos sistemas de informações do SUS, com as seguintes características: 1) Data Warehouse (DW) reunindo e integrando os dados dos sistemas do SUS; 2) processo de coleta, limpeza, integração e carga das bases de dados do SUS no DW; 3) componente para produção de informação gerencial; 4) metodologia para identificar o paciente em seus diversos atendimentos no Sistema Público de Saúde. Foram realizados diversos testes para avaliar a funcionalidade e a efetividade das ferramentas criadas, com ênfase em aplicações de cardiologia. Os resultados evidenciaram a efetividade das ferramentas nos aspectos mais complexos da gestão de informações para desenvolver conhecimentos, a partir das bases de dados do DATASUS

Managing medical images and clinical information: InCor's experience.

Patients usually get medical assistance in several clinics and hospitals during their lifetime, archiving vital information in a dispersed way. Clearly, a proper patient care should take into account that information in order to check for incompatibilities, avoid unnecessary exams, and get relevant clinical history. The Heart Institute (InCor) of São Paulo, Brazil, has been committed to the goal of integrating all exams and clinical information within the institution and other hospitals. Since InCor is one of the six institutes of the University of São Paulo Medical School and each institute has its own information system, exchanging information among the institutes is also a very important aspect that has been considered. In the last few years, a system for transmission, archiving, retrieval, processing, and visualization of medical images integrated with a hospital information system has been successfully created and constitutes the InCor's electronic patient record (EPR). This work describes the experience in the effort to develop a functional and comprehensive EPR, which includes laboratory exams, images (static, dynamic, and three dimensional), clinical reports, documents, and even real-time vital signals. A security policy based on a contextual role-based access control model was implemented to regulate user's access to EPR. Currently, more than 10 TB of digital imaging and communications in medicine (DICOM) images have been stored using the proposed architecture and the EPR stores daily more than 11 GB of integrated data. The proposed storage subsystem allows 6 months of visibility for rapid retrieval and more than two years for automatic retrieval using a jukebox. This paper addresses also a prototype for the integration of distributed and heterogeneous EPR.

O projeto Pacs: um sistema para visualizaçäo dinâmica e armazenamento de imagens de angiografia digital no Incor / Pacs project: a system to dynamic visualization and image storage of digital angiography in Incor

Este trabalho descreve a solução encontrada para um sistema PACS totalmente em padrão DICOM, esclarecendo as técnicas de desenvolvimento do sistema, métodos de compressão e visualização dinâmica para os estudos de cine-angiografia, o que permitirá, num futuro breve, a qualquer PC conectado à rede de computadores do InCor buscar, e visualizar sequências de imagens de forma segura, eficiente, com boa resolução espacial e temporal.