Biblioteca Digital do IPB
Publications Repository of the Polytechnic Institute of Bragança
Recent Submissions
Grape Winemaking By-Products: Current Valorization Strategies and Their Value as Source of Tannins with Applications in Food and Feed
Publication . Echave, Javier; González Pereira, Antía; Jorge, Ana O. S.; Barciela, Paula; Nogueira-Marques, Rafael; Yuksek, Ezgi N.; Oliveira, María B. P. P.; Barros, Lillian; Prieto, M. A.
Grape (Vitis vinifera L.) is one of the most extensively cultivated crops in temperate climates, with its primary fate being wine production, which is paired with a great generation of grape pomace (GP). GP contains a plethora of antioxidant phenolic compounds, being well-known for its high content of various tannins, liable for the astringency of this fruit. Winemaking produces a great mass of by-products that are rich in tannins. Grape seed (GSd) and pulp waste, as well as leaves and stems (GSt), are rich in condensed tannins (CTs), while its skin (GSk) contains more flavonols and phenolic acids. CTs are polymers of flavan-3-ols, and their antioxidant and anti-inflammatory properties are well-accounted for, being the subject of extensive research for various applications. CTs from the diverse fractions of grapefruit and grapevine share similar structures given their composition but diverge in their degree of polymerization, which can modulate their chemical interactions and may be present at around 30 to 80 mg/g, depending on the grape fraction. Thus, this prominent agroindustrial by-product, which is usually managed as raw animal feed or further fermented for liquor production, can be valorized as a source of tannins with high added value. The present review addresses current knowledge on tannin diversity in grapefruit and grapevine by-products, assessing the differences in composition, quantity, and degree of polymerization. Current knowledge of their reported bioactivities will be discussed, linking them to their current and potential applications in food and feed.
Binary classification of cardiac pathologies using deep learning: a PTB-XL dataset approach
Publication . Chaabani, Mohamed Khalil; Teixeira, João Paulo; Slim , Mohamed Aymen
Cardiovascular diseases, including myocardial infarction, remain among the leading causes of mortality worldwide. Timely and accurate diagnosis is critical for effective treatment but often requires labour-intensive manual analysis of clinical-grade electrocardiograms (ECGs). This dissertation proposes a novel deep learning-based approach for binary classification of cardiac pathologies, using the PTB-XL dataset. The final model architecture integrates EfficientNetB3 for spatial feature extraction and a Linformer block to capture long-range dependencies between ECG leads, the results prove its adaptability for ECG image classification tasks.
Extensive experimentation and iterative model development were conducted to reach the final design. Early trials involved exploring different hyperparameter tuning like Optuna and Adam optimizer and a wide range of hyperparameter configurations, including different learning rates, dropout rates, batch sizes, and numbers of Linformer layers. These experiments were critical in finding the optimal combination of parameters that balanced computational efficiency and model accuracy. Comprehensive details of these trials and evaluations are provided in the report.
The preprocessing pipeline involves selecting the ECGs and converting them to 11 images (aVR was excluded) each representing a lead, converting RGBA ECG images to RGB format and applying normalization to ensure compatibility with model input requirements. This preprocessing step addresses the unique format of the dataset and prepares it for high-performance neural network training.
Initial results from the finalized model architecture have demonstrated promising performance, achieving an AUC (Area Under the Curve) of 85.02% and a F1-score of 78.94%. The achieved results are comparable to recent state-of-the-art models reported on the PTB-XL dataset, which typically range between 85% and 95% AUC for similar binary classification tasks. These results indicate that the model's AUC of 85.02% is promising but on the edge of the current state-of-the-art. These findings indicate strong potential for the system to support clinical decision-making by automating the classification of ECG data. Ongoing research aims to extend the current binary classification framework to multi-class scenarios, further enhancing its clinical applicability. Additionally, efforts are being made to improve the model for faster inference times, enabling real-time ECG analysis and improving its feasibility for deployment in healthcare settings.
Electrospun polycaprolactone membranes functionalized with nanochitin for enhanced bioactivity in localized cancer photodynamic therapy
Publication . Costa, Sofia M.; Mattos, Bruno D.; Calhelha, Ricardo C.; Zhu, Ya; Lima, Eurico; Reis, Lucinda; Rojas, Orlando J.; Fangueiro, Raul; Ferreira, Diana P.
The encapsulation of photosensitizers (PSs) in electrospun membranes has emerged as a promising approach in photodynamic therapy (PDT) on tumor sites, overcoming the drawbacks associated with systemic administration. In this work, localized implants for cancer treatment using PDT were developed by incorporating EL-2 squaraine into poly(epsilon-caprolactone) (PCL) electrospun microfibers. The latter were coated with chitin nanocrystals (ChNC) by electrospraying, which may improve the biocompatibility and bioactivity of the developed membranes, potentially enhancing the clinical outcomes. The developed electrospun membranes were characterized by water contact angle, imaging, and spectroscopy techniques. The uniform encapsulation and distribution of EL-2 within the microfibers were confirmed while ChNC endowed the membranes with surface hydrophilicity. EL-2 alone displayed about 20 times more cytotoxicity after irradiation compared to the dark condition against HeLa cervical carcinoma cells. Meanwhile, the photodynamic action of PCL+EL-2/ChNC membranes promoted a significant inhibition of cancer cells' proliferation under irradiation, achieving 66.25 % of inhibition, compared to only 24.78 % in dark conditions, using the highest concentration of EL-2. Overall, this work introduces a disruptive strategy using electrospinning-electrospraying to design fibrous therapeutic platforms for cancer PDT, taking advantage of electrospun fibers unique features and the localized nature of photodynamic therapy.
Implementação de uma comunidade de energia renovável caso de estudo do Instituto Politécnico de Bragança
Publication . Baptista, Helder Ferreira; Leite, Vicente
A transição energética e a descentralização da produção elétrica são pilares fundamentais para o cumprimento dos compromissos ambientais assumidos por Portugal. Neste contexto, o presente trabalho desenvolve um estudo aprofundado sobre a caracterização, análise e gestão de energia no âmbito de uma Comunidade de Energia Renovável (CER) a ser implementada no Instituto Politécnico de Bragança (IPB), ao abrigo do Plano de Recuperação e Resiliência (PRR).
A dissertação baseia-se na análise detalhada de 14 Instalações de Utilização (IU), distribuídas por contratos de média tensão, baixa tensão especial e baixa tensão normal, com um consumo anual total de 2510 MWh (ano). Foram considerados dois cenários de consumo (anual e diurno no verão) para o dimensionamento de quatro Unidades de Produção para Autoconsumo (UPACs), com uma potência total instalada de 500 kW. O software PVSyst foi utilizado para simular a produção solar e otimizar a distribuição da potência entre os centros eletroprodutores, tendo-se escolhido o cenário anual como o mais equilibrado.
A proposta de gestão energética assenta num modelo de partilha hierárquica com coeficientes de partilha proporcionais ao consumo, priorizando os pontos de maior custo por kWh. Os resultados demonstram uma taxa média de autoconsumo de aproximadamente 86,5%, com variação sazonal dos excedentes. Adicionalmente, foi analisado o impacto da geração fotovoltaica na fatura de energia reativa, bem como as limitações estruturais dos edifícios para a instalação dos módulos fotovoltaicos.
A Multi-Dimensional Framework for Assessing the Societal Benefits of Collaborative R&I Projects Over Time
Publication . Brandão, Ana Sofia; Santos, José M.R.C.A.
This paper contributes to the ongoing discussion on assessing the actual societal benefits of collaborative research and innovation (R&I) projects, focusing specifically on Circular Bioeconomy (CBE) initiatives funded under European Interreg programs. Utilizing an abductive method aligned with a grounded theory approach, the study conducted a multiple case study of five cross-border CBE projects. Data from project leaders and secondary sources underwent inductive content analysis and were classified using the Triple Bottom Line (TBL) framework. Seven cross-cutting benefit categories emerged: capacity building, collaborative learning, community empowerment, networking, knowledge sharing, policy development, and sustainable business practices, identified as influencing results across TBL dimensions temporally. Findings reveal projects excel at generating short/medium-term outputs and outcomes strongly aligned with the social dimension, particularly through capacity building, collaborative learning, and knowledge sharing. Over time, long-term impacts demonstrate a more balanced distribution across all three TBL dimensions (social, environmental, and economic), indicating a trajectory towards broader benefits. Policy development and networking are emphasized as key drivers for achieving significant long-term, multi-dimensional impacts. This study introduces a novel, empirically grounded, multi-dimensional theoretical model. By inductively categorizing benefits and analyzing their temporal manifestation across TBL, it provides a practical framework for assessing comprehensive societal impact beyond conventional output metrics.