RESUMEN
OBJECTIVES: Machine learning (ML) is a powerful asset to support physicians in decision-making procedures, providing timely answers. However, ML for health systems can suffer from security attacks and privacy violations. This paper investigates studies of security and privacy in ML for health. METHODS: We examine attacks, defenses, and privacy-preserving strategies, discussing their challenges. We conducted the following research protocol: starting a manual search, defining the search string, removing duplicated papers, filtering papers by title and abstract, then their full texts, and analyzing their contributions, including strategies and challenges. Finally, we collected and discussed 40 papers on attacks, defense, and privacy. RESULTS: Our findings identified the most employed strategies for each domain. We found trends in attacks, including universal adversarial perturbation (UAPs), generative adversarial network (GAN)-based attacks, and DeepFakes to generate malicious examples. Trends in defense are adversarial training, GAN-based strategies, and out-of-distribution (OOD) to identify and mitigate adversarial examples (AE). We found privacy-preserving strategies such as federated learning (FL), differential privacy, and combinations of strategies to enhance the FL. Challenges in privacy comprehend the development of attacks that bypass fine-tuning, defenses to calibrate models to improve their robustness, and privacy methods to enhance the FL strategy. CONCLUSIONS: In conclusion, it is critical to explore security and privacy in ML for health, because it has grown risks and open vulnerabilities. Our study presents strategies and challenges to guide research to investigate issues about security and privacy in ML applied to health systems.
Asunto(s)
Médicos , Privacidad , Humanos , Aprendizaje AutomáticoRESUMEN
BACKGROUND: The image-based identification of distinct tissues within dermatological wounds enhances patients' care since it requires no intrusive evaluations. This manuscript presents an approach, we named QTDU, that combines deep learning models with superpixel-driven segmentation methods for assessing the quality of tissues from dermatological ulcers. METHOD: QTDU consists of a three-stage pipeline for the obtaining of ulcer segmentation, tissues' labeling, and wounded area quantification. We set up our approach by using a real and annotated set of dermatological ulcers for training several deep learning models to the identification of ulcered superpixels. RESULTS: Empirical evaluations on 179,572 superpixels divided into four classes showed QTDU accurately spot wounded tissues (AUC = 0.986, sensitivity = 0.97, and specificity = 0.974) and outperformed machine-learning approaches in up to 8.2% regarding F1-Score through fine-tuning of a ResNet-based model. Last, but not least, experimental evaluations also showed QTDU correctly quantified wounded tissue areas within a 0.089 Mean Absolute Error ratio. CONCLUSIONS: Results indicate QTDU effectiveness for both tissue segmentation and wounded area quantification tasks. When compared to existing machine-learning approaches, the combination of superpixels and deep learning models outperformed the competitors within strong significant levels.
Asunto(s)
Aprendizaje Profundo , Dermatología/métodos , Diagnóstico por Computador/métodos , Procesamiento de Imagen Asistido por Computador/métodos , Reconocimiento de Normas Patrones Automatizadas/métodos , Úlcera Cutánea/diagnóstico por imagen , Algoritmos , Área Bajo la Curva , Teorema de Bayes , Humanos , Aprendizaje Automático , Análisis de Componente Principal , Reproducibilidad de los Resultados , Sensibilidad y Especificidad , Máquina de Vectores de SoporteRESUMEN
This paper presents the extract-transform-and-load (ETL) process from the Electronic Patient Records (ePR) at the Heart Institute (InCor) to the OMOP Common Data Model (CDM) format. We describe the initial database characterization, relational source mappings, selection filters, data transformations and patient de-identification using the open-source OHDSI tools and SQL scripts. We evaluate the resulting InCor-CDM database by recreating the same patient cohort from a previous reference study (over the original data source) and comparing the cohorts' descriptive statistics and inclusion reports. The results exhibit that up to 91% of the reference patients were retrieved by our method from the ePR through InCor-CDM, with AUC=0.938. The results indicate that the method that we employed was able to produce a new database that was both consistent with the original data and in accordance to the OMOP CDM standard.
Asunto(s)
Registros Electrónicos de Salud , Almacenamiento y Recuperación de la Información , Estudios de Cohortes , Bases de Datos Factuales , Atención a la Salud , HumanosRESUMEN
In this paper we address the "skull-stripping" problem in 3D MR images. We propose a new method that employs an efficient and unique histogram analysis. A fundamental component of this analysis is an algorithm for partitioning a histogram based on the position of the maximum deviation from a Gaussian fit. In our experiments we use a comprehensive image database, including both synthetic and real MRI, and compare our method with other two well-known methods, namely BSE and BET. For all datasets we achieved superior results. Our method is also highly independent of parameter tuning and very robust across considerable variations of noise ratio.
Asunto(s)
Imagen por Resonancia Magnética/métodos , Cráneo/anatomía & histología , Algoritmos , Sistemas de Administración de Bases de Datos , HumanosRESUMEN
En el presente artículo se propone una técnica de caracterización de imágenes médicas aplicando una metodología de multi-resolución, con la finalidad de facilitar su respectiva indexación dentro de una estructura multi-dimensional. Una de las características de las imágenes médicas, es la de presentar cambios de color en tonos de gris dentro de regiones localizadas de la imagen. Hasta ahora no se tiene una técnica adecuada que haga posible extraer estas regiones mediante un completo procesamiento automático de las imágenes. Una estrategia para abordar este problema consiste en la generación de vectores de características basados en las transformadas de wavelet, estas características extraídas generaran un vector de características que se va a constituir en la identificación de la imagen. En la presente propuesta, el sistema extrae las características más relevantes de la imagen, calcula la distancia entre una imagen de consulta y las que seencuentran en el banco de datos, y recupera las n imágenes más similares. El enfoque presentado estábasado en la aplicación de los filtros de las wavelets de Daubechies 4 sobre las características globales de la imagen para luego así facilitar la indexación dentro de una estructura multi-dimensional. El objetivo del presente trabajo es destacar la utilidad de la transformada de wavelet en la caracterización de imágenes médicas y su utilidad dentro de una apropiada técnica de indexación de las mismas.
Asunto(s)
Bases de Datos como Asunto , Procesamiento de Imagen Asistido por ComputadorRESUMEN
This paper presents a new Picture Archiving and Communication System (PACS), called cbPACS, which has content-based image retrieval capabilities. The cbPACS answers range and k-nearest- neighbor similarity queries, employing a relational database manager extended to support images. The images are compared through their features, which are extracted by an image-processing module and stored in the extended relational database. The database extensions were developed aiming at efficiently answering similarity queries by taking advantage of specialized indexing methods. The main concept supporting the extensions is the definition, inside the relational manager, of distance functions based on features extracted from the images. An extension to the SQL language enables the construction of an interpreter that intercepts the extended commands and translates them to standard SQL, allowing any relational database server to be used. By now, the system implemented works on features based on color distribution of the images through normalized histograms as well as metric histograms. Metric histograms are invariant regarding scale, translation and rotation of images and also to brightness transformations. The cbPACS is prepared to integrate new image features, based on texture and shape of the main objects in the image.
Asunto(s)
Almacenamiento y Recuperación de la Información , Sistemas de Información Radiológica , Lenguajes de ProgramaciónRESUMEN
Este trabalho apresenta a construção de um Gerenciador de Bases de Dados para a armazenagem e recuperação de imagens, que permite a formulação de consultas através de seu conteúdo. As abordagens usuais estabelecem a utilização de ícones e atributos textuais para a recuperação das imagens. Neste trabalho é apresentada uma modelagem para o "tipo de dado imagem", que permite a construção de sistemas expansíveis, onde novos algoritmos de processamento de imagens podem ser incluídos, através de sua descrição num meta-esquema de dados. As operações executadas pelos algoritmos modelados podem ser acessadas via a linguagem de consulta do Gerenciador.