Helder Coelho
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Helder Coelho has a total of 112 co-authors. Here are the top co-authors by number of shared publications:
José Carlos Cotta
Laboratório Nacional de Engenharia Civil
18 shared publications- Luis Antunes
Faculdade de Ciências LASIGE Universidade de Lisboa Portugal
16 shared publications - Fernando Lopes
National Laboratory of Energy and Geology, 1649-038 Lisbon, Portugal
14 shared publications - Paulo Trigo
Instituto Superior de Engenharia de Lisboa
14 shared publications - Nuno David
ISCTE-IUL Instituto Universitário de Lisboa
8 shared publications - Augusto Queiroz Novais
DMS, Instituto Nacional de Engenharia, Tecnologia e Inovação, Est. do Paço do Lumiar, 1649-038 Lisboa, Portugal, and CEG-IST, Instituto Superior Técnico, Univ. Técnica de Lisboa, Av. Rovisco Pais, 1049-101 Lisboa, Portugal
7 shared publications
Latest Publications
- 18 October 2025
- Published by Springer Nature
- p. 167-178
Agenda-Setting in the Brazilian Presidential Elections 2018 on Twitter
- 18 October 2025
- Published by Springer Nature
- p. 167-178
Agenda-Setting in the Brazilian Presidential Elections 2018 on Twitter
- Open Access
- 27 December 2023
- Published by Springer Nature
- p. 169-185
Metacognition, Accountability and Legal Personhood of AI
- 27 December 2023
- Published by Springer Nature
- p. 169-185
Open AccessMetacognition, Accountability and Legal Personhood of AI
- 14 April 2023
- Published by Georg Thieme Verlag KG in Endoscopy
- Vol. 55 (S 02) , S148
Development of a combined deep learning model for automatic detection of multiple gastrointestinal lesions in device-assisted enteroscopy using convolutional neural networks
Aims Device-assisted enteroscopy (DAE) allows deep exploration of the gastrointestinal (GI) tract, combining its diagnostic ability and the application of endoscopic therapy. Recently, a large number of studies have been published regarding the application of artificial intelligence (AI) to different endoscopic modalities. We aimed to develop a combined panenteric convolutional neural network (CNN) to identify multiple gastrointestinal lesions using DAE images ([Table 1]). Methods We developed a combined CNN-based system based on 250 DAE exams performed at a single center. A total of 12870 images from different segments of the GI tract (stomach, small bowel and colon) were included, 2668images containing protruding lesions, 1547 hematic residues, 1450 angioectasias, 633 containing ulcers/erosions. The remaining images showed normal mucosa (n=6139) and other findings not classified elsewhere (e.g. xanthomas, stenosis and diverticula, n=433). Training and validation datasets were built using 80% and 20% of the full dataset, respectively [1] [2] [3]. Results The model had na overall sensitivity of 96.2%, specificity 95.0%, positive and negative predictive values of 95.6% and 95.7%, respectively, and an overall accuracy of 95.6%. The AUC was 0.99. The results for specific lesions are summarized in table 1. Conclusions The authors have developed a pioneer combined deep learning system for automatic detection of a wide range of lesions across the different segments of the gastrointestinal tract. The development and application of these technologies DAE techniques may complement current existing panenteric techniques (e.g. capsule endoscopy), providing more accurate diagnosis and directed treatments. Article published online: 14 April 2023 © 2023. European Society of Gastrointestinal Endoscopy. All rights reserved. Georg Thieme Verlag KG Rüdigerstraße 14, 70469 Stuttgart, Germany
- 14 April 2023
- Published by Georg Thieme Verlag KG in Endoscopy
- Vol. 55 (S 02) , S148
Development of a combined deep learning model for automatic detection of multiple gastrointestinal lesions in device-assisted enteroscopy using convolutional neural networks
Aims Device-assisted enteroscopy (DAE) allows deep exploration of the gastrointestinal (GI) tract, combining its diagnostic ability and the application of endoscopic therapy. Recently, a large number of studies have been published regarding the application of artificial intelligence (AI) to different endoscopic modalities. We aimed to develop a combined panenteric convolutional neural network (CNN) to identify multiple gastrointestinal lesions using DAE images ([Table 1]). Methods We developed a combined CNN-based system based on 250 DAE exams performed at a single center. A total of 12870 images from different segments of the GI tract (stomach, small bowel and colon) were included, 2668images containing protruding lesions, 1547 hematic residues, 1450 angioectasias, 633 containing ulcers/erosions. The remaining images showed normal mucosa (n=6139) and other findings not classified elsewhere (e.g. xanthomas, stenosis and diverticula, n=433). Training and validation datasets were built using 80% and 20% of the full dataset, respectively [1] [2] [3]. Results The model had na overall sensitivity of 96.2%, specificity 95.0%, positive and negative predictive values of 95.6% and 95.7%, respectively, and an overall accuracy of 95.6%. The AUC was 0.99. The results for specific lesions are summarized in table 1. Conclusions The authors have developed a pioneer combined deep learning system for automatic detection of a wide range of lesions across the different segments of the gastrointestinal tract. The development and application of these technologies DAE techniques may complement current existing panenteric techniques (e.g. capsule endoscopy), providing more accurate diagnosis and directed treatments. Article published online: 14 April 2023 © 2023. European Society of Gastrointestinal Endoscopy. All rights reserved. Georg Thieme Verlag KG Rüdigerstraße 14, 70469 Stuttgart, Germany
- 11 January 2023
- Published by Springer Nature
- p. 147-155
Interdisciplinary Practice in Education
- 13 September 2022
- Published by Springer Nature
- p. 512-521
On Developing Ethical AI
- 13 September 2022
- Published by Springer Nature
- p. 751-763
Bringing Underused Learning Objects to the Light: A Multi-agent Based Approach
- 13 September 2022
- Published by Springer Nature
- p. 751-763
Bringing Underused Learning Objects to the Light: A Multi-agent Based Approach
- 23 April 2021
- Published by Springer Nature
- p. 396-401
A Proposal for Artificial Moral Pedagogical Agents
- 23 April 2021
- Published by Springer Nature
- p. 396-401
A Proposal for Artificial Moral Pedagogical Agents
- Open Access
SensAI+Expanse Adaptation on Human Behaviour Towards Emotional Valence Prediction
An agent, artificial or human, must be continuously adjusting its behaviour in order to thrive in a more or less demanding environment. An artificial agent with the ability to predict human emotional valence in a geospatial and temporal context requires proper adaptation to its mobile device environment with resource consumption strict restrictions (e.g., power from battery). The developed distributed system includes a mobile device embodied agent (SensAI) plus Cloud-expanded (Expanse) cognition and memory resources. The system is designed with several adaptive mechanisms in a best effort for the agent to cope with its interacting humans and to be resilient on collecting data for machine learning towards prediction. These mechanisms encompass homeostatic-like adjustments such as auto recovering from an unexpected failure in the mobile device, forgetting repeated data to save local memory, adjusting actions to a proper moment (e.g., notify only when human is interacting), and the Expanse complementary learning algorithms' parameters with auto adjustments. Regarding emotional valence prediction performance, results from a comparison study between state-of-the-art algorithms revealed Extreme Gradient Boosting on average the best model for prediction with efficient energy use, and explainable using feature importance inspection. Therefore, this work contributes with a smartphone sensing-based system, distributed in the Cloud, robust to unexpected behaviours from humans and the environment, able to predict emotional valence states with very good performance.
Open AccessSensAI+Expanse Adaptation on Human Behaviour Towards Emotional Valence Prediction
An agent, artificial or human, must be continuously adjusting its behaviour in order to thrive in a more or less demanding environment. An artificial agent with the ability to predict human emotional valence in a geospatial and temporal context requires proper adaptation to its mobile device environment with resource consumption strict restrictions (e.g., power from battery). The developed distributed system includes a mobile device embodied agent (SensAI) plus Cloud-expanded (Expanse) cognition and memory resources. The system is designed with several adaptive mechanisms in a best effort for the agent to cope with its interacting humans and to be resilient on collecting data for machine learning towards prediction. These mechanisms encompass homeostatic-like adjustments such as auto recovering from an unexpected failure in the mobile device, forgetting repeated data to save local memory, adjusting actions to a proper moment (e.g., notify only when human is interacting), and the Expanse complementary learning algorithms' parameters with auto adjustments. Regarding emotional valence prediction performance, results from a comparison study between state-of-the-art algorithms revealed Extreme Gradient Boosting on average the best model for prediction with efficient energy use, and explainable using feature importance inspection. Therefore, this work contributes with a smartphone sensing-based system, distributed in the Cloud, robust to unexpected behaviours from humans and the environment, able to predict emotional valence states with very good performance.
- 30 August 2019
- Published by Springer Nature
- p. 633-641
Reflections on Social Simulation and Complexity
- Open Access
- 30 December 2018
- Published by ICETA in Revista de Ciência Elementar
- Vol. 6 (4)
Simulação Social
A Simulação Social refere-se ao uso de métodos computacionais e analíticos, baseados na Ciência da Complexidade e na Ciência dos Dados e, ainda, no rigor matemático para melhorar a compreensão dos sistemas sociais (por exemplo, alguns problemas nacionais, como a mobilidade urbana e os transportes, a crise financeira global, a saúde pública, ou a educação).
- 30 December 2018
- Published by ICETA in Revista de Ciência Elementar
- Vol. 6 (4)
Open AccessSimulação Social
A Simulação Social refere-se ao uso de métodos computacionais e analíticos, baseados na Ciência da Complexidade e na Ciência dos Dados e, ainda, no rigor matemático para melhorar a compreensão dos sistemas sociais (por exemplo, alguns problemas nacionais, como a mobilidade urbana e os transportes, a crise financeira global, a saúde pública, ou a educação).