CIIC - Artigos em Revistas com Peer Review
Permanent URI for this collection
Browse
Browsing CIIC - Artigos em Revistas com Peer Review by Title
Now showing 1 - 10 of 38
Results Per Page
Sort Options
- 2ARTs: A Platform for Exercise Prescriptions in Cardiac Recovery PatientsPublication . Pereira, Andreia; Martinho, Ricardo; Pinto, Rui; Rijo, Rui; Grilo, CarlosDue to limited access, increasing costs and an ageing population, the global healthcare system faces significant coverage problems that call for innovative approaches. Health professionals are actively seeking alternative methods to provide care to an increasingly needy population, without increasing human effort and associated costs. eHealth platforms, which use technology to provide patient care, are emerging as transformative solutions for addressing these problems. This study is centered on the demand for a Decision Support System (DSS) in cardiology to enable doctors to prescribe individualized care inside Cardiac Rehabilitation Programmes (CRPs). The 2ARTs project’s main objective is to include a cardiac rehabilitation platform with a DSS within the hospital infrastructure. This DSS uses models to classify patients into different groups, delivering crucial information to assist with decisions regarding treatment. Regarding the DSS, Principal Component Analysis (PCA) emerged as a standout technique for dimensionality reduction, due to its interoperability with clustering algorithms and superior evaluation metrics. The most appropriate clustering technique was determined to be the K-means algorithm, which was supported by the experts analysis. In accordance with the goals of the 2ARTs project, this integration of PCA and K-means provides meaningful insights that improve reasoned decision-making.
- 3D fast convex-hull-based evolutionary multiobjective optimization algorithmPublication . Zhao, Jiaqi; Jiao, Licheng; Liu, Fang; Basto-Fernandes, Vitor; Yevseyeva, Iryna; Xia, Shixiong; Emmerich, Michael T.M.The receiver operating characteristic (ROC) and detection error tradeoff (DET) curves have been widely used in the machine learning community to analyze the performance of classifiers. The area (or volume) under the convex hull has been used as a scalar indicator for the performance of a set of classifiers in ROC and DET space. Recently, 3D convex-hull-based evolutionary multiobjective optimization algorithm (3DCH-EMOA) has been proposed to maximize the volume of convex hull for binary classification combined with parsimony and three-way classification problems. However, 3DCH-EMOA revealed high consumption of computational resources due to redundant convex hull calculations and a frequent execution of nondominated sorting. In this paper, we introduce incremental convex hull calculation and a fast replacement for non-dominated sorting. While achieving the same high quality results, the computational effort of 3DCH-EMOA can be reduced by orders of magnitude. The average time complexity of 3DCH-EMOA in each generation is reduced from to per iteration, where n is the population size. Six test function problems are used to test the performance of the newly proposed method, and the algorithms are compared to several state-of-the-art algorithms, including NSGA-III, RVEA, etc., which were not compared to 3DCH-EMOA before. Experimental results show that the new version of the algorithm (3DFCH-EMOA) can speed up 3DCH-EMOA for about 30 times for a typical population size of 300 without reducing the performance of the method. Besides, the proposed algorithm is applied for neural networks pruning, and several UCI datasets are used to test the performance.
- Analyzing TikTok from a Digital Forensics PerspectivePublication . Domingues, Patricio; Nogueira, Ruben; Francisco, José Carlos; Frade, MiguelTikTok is a major hit in the digital mobile world, quickly reaching the top 10 installed applications for the two main mobile OS, iOS and Android. This paper studies Android's TikTok application from a digital forensic perspective, analyzing the digital forensic artifacts that can be retrieved on a post mortem analysis and their associations with operations performed by the user. The paper also presents FAMA (Forensic Analysis for Mobile Apps), an extensible framework for the forensic software Autopsy, and FAMA's TikTok module that collects, analyzes, and reports on the main digital forensic artifacts of TikTok's Android application. The most relevant digital artifacts of TikTok include messages exchanged between TikTok so-called ``friends'', parts of the email/phone number of registered users, data about devices, and transactions with TikTok's virtual currency. One of the results of this research is the set of forensic traces left by users' transactions with TikTok's in-app virtual currency. Another result is the detection of patterns that exist in TikTok's integer IDs, allowing to quickly link any 64-bit TikTok's integer ID to the type of resources -- user, device, video, etc. -- that it represents.
- Applying deep learning to real-time UAV-based forest monitoring: Leveraging multi-sensor imagery for improved resultsPublication . Marques, Tomás; Carreira, Samuel; Miragaia, Rolando; Ramos, João; Pereira, AntónioRising global fire incidents necessitate effective solutions, with forest surveillance emerging as a crucial strategy. This paper proposes a complete solution using technology that integrates visible and infrared spectrum images through Unmanned Aerial Vehicles (UAVs) for enhanced detection of people and vehicles in forest environments. Unlike existing computer vision models relying on single-sensor imagery, this approach overcomes limitations posed by limited spectrum coverage, particularly addressing challenges in low-light conditions, fog, or smoke. The developed 4-channel model uses both types of images to take advantage of the strengths of each one simultaneously. This article presents the development and implementation of a solution for forest monitoring ranging from the transmission of images captured by a UAV to their analysis with an object detection model without human intervention. This model consists of a new version of the YOLOv5 (You Only Look Once) architecture. After the model analyzes the images, the results can be observed on a web platform on any device, anywhere in the world. For the model training, a dataset with thermal and visible images from the aerial perspective was captured with a UAV. From the development of this proposal, a new 4- channel model was created, presenting a substantial increase in precision and mAP (Mean Average Precision) metrics compared to traditional SOTA (state-of-the-art) models that only make use of red, green, and blue (RGB) images. Allied with the increase in precision, we confirmed the hypothesis that our model would perform better in conditions unfavorable to RGB images, identifying objects in situations with low light and reduced visibility with partial occlusions. With the model’s training using our dataset, we observed a significant increase in the model’s performance for images in the aerial perspective. This study introduces a modular system architecture featuring key modules: multisensor image capture, transmission, processing, analysis, and results presentation. Powered by an innovative object detection deep-learning model, these components collaborate to enable real-time, efficient, and distributed forest monitoring across diverse environments.
- Care4Value: medição de valor em saúde em Unidades de Cuidados Continuados IntegradosPublication . Reis, Catarina I.; Maximiano, Marisa; Ferreira, Pedro Henrique; Querido, Ana; Sargento, Ana Lúcia Marto; Carvalho, Henrique; Leal, Susana Cristina Henriques; Oliveira, Sandra Margarida Bernardes deObjetivo: Desenvolver uma plataforma digital para a otimização do processo de coleta de dados de escalas clínicas e monitoramento desses dados com vista à medição do valor em saúde. Métodos: Por meio de uma metodologia de investigação-ação, o desenvolvimento da plataforma incluiu abordagens qualitativas e quantitativas, em três fases: grupos focais com uma equipe multidisciplinar de investigadores e profi ssionais de saúde da UCCI do estudo-piloto; análise dos dados clínicos em formato de pré-teste de uma amostra de 21 usuários da UCCI para categorizar diferentes graus de complexidade; e, análise de informação fi nanceira, aos custos operacionais da UCCI, relativa ao momento de permanência dos mesmos 21 usuários. O desenvolvimento iterativo e incremental da plataforma permitiu coletar feedback dos usuários como forma de melhoria. Resultados: A plataforma inclui 3 módulos: aplicativo móvel; dashboard; e módulo de importação. A plataforma centraliza os dados coletados e disponibiliza-os por meio de um dashboard. Os dados são coletados por aplicativo móvel e/ou por um módulo de importação que consome dados de sistemas clínicos existentes. Conclusão: O aplicativo móvel está apto a ser utilizado por profi ssionais de saúde e cuidadores, e o dashboard apresenta informações de acompanhamento clínico dos usuários e monitoramento dos seus ganhos em saúde.
- Convolutional-LSTM networks and generalization in forecasting of household photovoltaic generationPublication . Costa, Rogério Luís de C.Solar panels can generate energy to meet almost all of the energy needs of a house. Batteries store energy generated during daylight hours for future use. Also, it may be possible to sell extra electricity back to distribution companies. However, the efficiency of photovoltaic systems varies according to several factors, such as the solar exposition at ground levels, atmospheric temperature, and relative humidity, and predicting the energy generated by such a system is not easy. This work is on the use of deep learning to predict the generation of photovoltaic energy by residential systems. We use real-world data to evaluate the performance of LSTM, Convolutional, and hybrid Convolutional-LSTM networks in predicting photovoltaic power generation at different forecasting horizons. We also assess the generalizability of the solutions, evaluating the use of models trained with data aggregated by geographic areas to predict the energy generation by individual systems. We compare the performance of deep networks with Prophet in terms of MAE, RMSE, and NRMSE, and in most cases, Convolutional and Convolutional-LSTM networks achieve the best results. Using models trained with region-based data to predict the power generation of individual systems is confirmed to be a promising approach.
- Cybersecurity risk analysis in healthcare institutionsPublication . Nunes, P; Antunes, M; Silva, CIntroduction The growing digitization of businesses and its increasing dependence on Internet infrastructure has boosted the concerns related to data privacy and confidentiality. Healthcare institutions have been challenged with specific issues, namely the sensitivity of data, the specificity of networked equipment and the average information technology skills held by of healthcare professionals in Portugal.
- Decrypting messages: Extracting digital evidence from signal desktop for windowsPublication . Paulino, Gonçalo; Negrão, Miguel; Frade, Miguel; Domingues, PatrícioWith growing concerns over the security and privacy of personal conversations, end-to-end encrypted instant messaging applications have become a key focus of forensic research. This study presents a detailed methodology along with an automated Python script for decrypting and analyzing forensic artifacts from Signal Desktop for Windows. The methodology is divided into two phases: i) decryption of locally stored data and ii) analysis and documentation of forensic artifacts. To ensure data integrity, the proposed approach enables retrieval without launching Signal Desktop, preventing potential alterations. Additionally, a reporting module organizes extracted data for forensic investigators, enhancing usability. Our approach is effective in extracting and analyzing encrypted Signal artifacts, providing a reliable method for forensic investigations.
- A Digital Forensic View of Windows 10 NotificationsPublication . Domingues, Patricio; Andrade, Luís; Frade, MiguelWindows Push Notifications (WPN) is a relevant part of Windows 10 interaction with the user. It is comprised of badges, tiles and toasts. Important and meaningful data can be conveyed by notifications, namely by so-called toasts that can popup with information regarding a new incoming email or a recent message from a social network. In this paper, we analyze the Windows 10 Notification systems from a digital forensic perspective, focusing on the main forensic artifacts conveyed by WPN. We also briefly analyze Windows 11 first release’s WPN system, observing that internal data structures are practically identical to Windows 10. We provide an open source Python 3 command line application to parse and extract data from the Windows Push Notification SQLite3 database, and a Jython module that allows the well-known Autopsy digital forensic software to interact with the application and thus to also parse and process Windows Push Notifications forensic artifacts. From our study, we observe that forensic data provided by WPN are scarce, although they still need to be considered, namely if traditional Windows forensic artifacts are not available. Furthermore, toasts are clearly WPN’s most relevant source of forensic data.
- Driving Behavior Classification Using a ConvLSTMPublication . Pingo, Alberto; Castro, João; Loureiro, Paulo; Mendes, Silvio; Bernardino, Anabela; Miragaia, Rolando; Husyeva, IrynaThis work explores the classification of driving behaviors using a hybrid deep learning model that combines Convolutional Neural Networks (CNNs) with Long Short-Term Memory (LSTM) networks (ConvLSTM). Sensor data are collected from a smartphone application and undergo a preprocessing pipeline, including data normalization, labeling, and feature extraction, to enhance the model’s performance. By capturing temporal and spatial dependencies within driving patterns, the proposed ConvLSTM model effectively differentiates between normal and aggressive driving behaviors. The model is trained and evaluated against traditional stacked LSTM and Bidirectional LSTM (BiLSTM) architectures, demonstrating superior accuracy and robustness. Experimental results confirm that the preprocessing techniques improve classification performance, ensuring high reliability in driving behavior recognition. The novelty of this work lies in a simple data preprocessing methodology combined with the specific application scenario. By enhancing data quality before feeding it into the AI model, we improve classification accuracy and robustness. The proposed framework not only optimizes model performance but also demonstrates practical feasibility, making it a strong candidate for real-world deployment.