A5017050304- ECG Monitoring and Analysis
- Machine Learning in Healthcare
- Artificial Intelligence in Healthcare
- Time Series Analysis and Forecasting
- Cardiovascular Function and Risk Factors
- Heart Rate Variability and Autonomic Control
- Cardiac Imaging and Diagnostics
- Complex Systems and Time Series Analysis
- Explainable Artificial Intelligence (XAI)
- Cardiac electrophysiology and arrhythmias
- Acute Myocardial Infarction Research
- Anomaly Detection Techniques and Applications
- Non-Invasive Vital Sign Monitoring
- Blood Pressure and Hypertension Studies
- Diabetes, Cardiovascular Risks, and Lipoproteins
- EEG and Brain-Computer Interfaces
- Health Systems, Economic Evaluations, Quality of Life
- Artificial Intelligence in Healthcare and Education
- Anesthesia and Pain Management
- Spectroscopy and Chemometric Analyses
- Imbalanced Data Classification Techniques
- Cardiac Health and Mental Health
- Heart Failure Treatment and Management
- Fault Detection and Control Systems
- Phonocardiography and Auscultation Techniques
Universidade Estadual do Ceará
2025
Polytechnic Institute of Coimbra
2015-2024
University of Coimbra
2012-2024
Novo Nordisk (Denmark)
2024
Centro Hospitalar de Lisboa Central
2023
Centro De Medicina Veterinária Anjos De Assis
2023
Institute of Engineering
2021-2022
Escola Superior de Tecnologia da Saúde de Coimbra
2016
Institute for Systems Engineering and Computers
2008-2012
Instituto Politécnico Nacional
2009-2011
Cardiovascular disease has a huge impact on health care services, originating unsustainable costs at clinical, social, and economic levels. In this context, patients' risk stratification tools are central to support clinical decisions contributing the implementation of effective preventive care. Although useful, these present some limitations, in particular, lack performance as well impossibility consider new factors potentially important prognosis severe cardiac events. Moreover, actual use...
Searching for similarity between time series plays an important role when large amounts of information need to be clustered integrate intelligent supported personal health care diagnosis systems. The performance classification, clustering and disease prediction are influenced by the prior stage where is performed. Physiologic signals vary even within same patient, so analysis their possible variation without affecting future accuracy hereby addressed. Commonly employed methods measuring were...
This paper aims to assess the predictive value of physiological data daily collected in a telemonitoring study early detection heart failure (HF) decompensation events. The main hypothesis is that time series with similar progression (trends) may have prognostic future clinical states (decompensation or normal condition). strategy composed two steps: trend similarity analysis and procedure. scheme combines Haar wavelet decomposition, which signals are represented as linear combinations set...
Summary The presence of body posture changes among patients with temporomandibular disorders ( TMD ) has been a controversial topic in dentistry. Based on that, the aim this study was to assess postural features pain‐free subjects internal derangement joint TMJ ), viz. disc displacement, when compared normal position. A total 21 unilateral, displacement DD and without any signs symptoms were assessed for by means posturographic evaluation several segments balance reactions through centre...
IcoSema is being developed as a subcutaneous once-weekly fixed-ratio combination of the basal insulin icodec and glucagon-like peptide-1 receptor agonist semaglutide. This study investigated pharmacokinetics semaglutide in versus separate administration each component individuals with type 2 diabetes mellitus (T2DM). In randomised, double-blind, three-period crossover study, 31 T2DM (18–64 years, body weight 80–120 kg, glycosylated haemoglobin 6.0–8.5%) received single injections (175 U...
This paper presents a generic methodology for time series prediction, based on wavelet decomposition/ reconstruction technique, together with feedforward neural networks structure. The proposed combines the flexibility and learning abilities of compact description signals, inherent to wavelets. In first phase decomposition signal is performed, providing small number coefficients that summarizes evolution dynamics. prediction problem then effectively addressed by means model, previously...
Ventricular arrhythmias, especially tachycardia and fibrillation are one of the main causes sudden cardiac death. Therefore, development methodologies, enable to detect their occurrence characterize time evolution, is fundamental importance. This work proposes a non-linear dynamic signal processing approach address problem. Based on phase space reconstruction electrocardiogram (ECG), some features extracted for each ECG window. Features from current previous windows provided neural network...
The cardioRisk project addresses the development of personalized risk assessment tools for patients who have been admitted to hospital with acute myocardial infarction. Although there are models available that assess short-term death/new events such patients, these were established in circumstances do not take into account present clinical interventions and, some cases, factors used by easily practice. integration existing (applied clinician's daily practice) data-driven knowledge discovery...
The present work aims to an innovative measure able efficiently evaluate the similarity between two physiological time series. proposed methodology combines Haar wavelet decomposition, in which signals are represented as linear combinations of a set orthogonal basis, with Karhunen-Loève transform, that allows for optimal reduction basis. is based on Euclidean distance, but indirectly calculated through combination coefficients both Moreover, iterative scheme computing referred significantly...
This work proposes a wavelet decomposition based scheme to estimate the evolution trend of physiological time series. The does not involve explicit development model and is essentially supported on hypothesis that future biosignal can be estimated from similar historic patterns. strategy considers an a-trous decomposition, where most representative trends are extracted Then, set distance-based measures able assess prediction likelihood each trend, introduced. From these through optimization...
Clinical guidelines recommend the use of cardiovascular risk assessment tools (risk scores) to predict events such as death, since these scores can aid clinical decision-making and thereby reduce social economic costs disease (CVD). However, despite their importance, present important weaknesses that diminish reliability in contexts. This study presents a new framework, based on current tools, aims minimize limitations. Appropriate application combination existing knowledge is main focus...
This work proposes a framework for telehealth streams analysis, founded on pattern recognition technique that evaluates the similarity between multi-sensorial biosignals. The strategy combines Haar wavelet with Karhunen-Loève transforms to describe biosignals by means of reduced set parameters. These, reflect dynamic behavior biosignals, can support detection relevant clinical conditions. Moreover, simplicity and fast execution proposed approach allow its application in real-time operation,...
Reduced ejection fraction (EF), possibly induced/mediated by autonomic abnormal activation, is one of the most powerful predictors adverse outcome after acute myocardial infarction (MI). A deep understanding correlation between autonomous functionality and left ventricular performance in these patients therefore paramount importance. The function reflected cardiac activity and, specifically, heart rate variability (HRV) signal. Given nonlinearity, growing interest being manifested towards...
Two innovative CVD event risk assessment strategies were developed in the scope of HeartCycle project: i) combination individual tools; ii) personalization based on grouping patients. These approaches aimed to defeat some major limitations tools currently applied daily clinical practice, namely to: improve prediction performance when comparing it one achieved by current consider available knowledge provided other iii) cope with missing factors; iv) incorporate additional knowledge. different...
This work describes a MATLAB tool developed in the context of didactic application toolbox that implements some advanced optimization and decision support methodologies, intended for use by undergraduate students. The particular module addressed this paper solves linear goal programming problems, not only analytically, but also graphically if number variables is less than three. Although was designed to students study specific course unit area, once appropriately tested it can be made...