Juan Antonio Recio Garcia
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Juan Antonio Recio Garcia has a total of 94 co-authors. Here are the top co-authors by number of shared publications:
- Belén Díaz-Agudo
Universidad Complutense de Madrid
43 shared publications - Marta Caro-Martínez
Universidad Complutense de Madrid
14 shared publications - Pedro Antonio Gonzalez-Calero
Universidad Complutense de Madrid, Madrid, Spain
11 shared publications - Jesus M. Darias
Universidad Complutense de Madrid
7 shared publications - Guillermo Jimenez-Diaz
Universidad Complutense de Madrid
5 shared publications - Jose Luis Jorro-Aragoneses
Universidad Autónoma de Madrid
5 shared publications
Latest Publications
- 6 November 2026
- Published by Springer Nature
- p. 332-344
Cellular Automata, $(\mu + \lambda )$ Evolutionary Strategies and Morphogenesis: A CBR Approach- 6 November 2026
- Published by Springer Nature
- p. 332-344
Cellular Automata, $(\mu + \lambda )$ Evolutionary Strategies and Morphogenesis: A CBR Approach- 1 August 2026
- Published by Springer Nature
- p. 521-536
Knowledge-Guided Generative Adaptation in Case-Based Reasoning for Automatic Art Exhibition Generation
- 1 August 2026
- Published by Springer Nature
- p. 521-536
Knowledge-Guided Generative Adaptation in Case-Based Reasoning for Automatic Art Exhibition Generation
- Open Access
- 13 July 2026
- Published by Universidad Internacional de La Rioja in International Journal of Interactive Multimedia and Artificial Intelligence
- Vol. 10 (1) , e6503
XAIoT: a Conceptual Model for XAI Solutions on IoT
Artificial intelligence (AI) systems are part of our current world, helping us to perform a wide variety of tasks. In parallel, eXplainable Artificial Intelligence (XAI), which obtains explanations to help users understand AI systems, has increased in significance in the last few years. This relevance comes from users’ necessity to trust new complex AI models. However, AI is not confined to computers but is also applied to other devices to use current internet connectivity to make our lives easier. The already well-known Internet of Things (IoT) refers to this digitally connected universe of smart devices. Hence, XAI should also be applied to the IoT, although there is a gap in the literature concerning conceptualizing explanation methods applied to AI models on the IoT. In this work, we propose a conceptual model named XAIoT (eXplainable Artificial Intelligence of Things) that structures and formalizes the main features of such systems. The ultimate goal is to provide a framework that guides the selection of the most suitable XAI solution for a given intelligent IoT application. This conceptual model is validated by classifying up to 100 approaches obtained from a systematic review of the state-of-the-art and formalized in an ontology to guide XAIoT system design.
- 13 July 2026
- Published by Universidad Internacional de La Rioja in International Journal of Interactive Multimedia and Artificial Intelligence
- Vol. 10 (1) , e6503
Open AccessXAIoT: a Conceptual Model for XAI Solutions on IoT
Artificial intelligence (AI) systems are part of our current world, helping us to perform a wide variety of tasks. In parallel, eXplainable Artificial Intelligence (XAI), which obtains explanations to help users understand AI systems, has increased in significance in the last few years. This relevance comes from users’ necessity to trust new complex AI models. However, AI is not confined to computers but is also applied to other devices to use current internet connectivity to make our lives easier. The already well-known Internet of Things (IoT) refers to this digitally connected universe of smart devices. Hence, XAI should also be applied to the IoT, although there is a gap in the literature concerning conceptualizing explanation methods applied to AI models on the IoT. In this work, we propose a conceptual model named XAIoT (eXplainable Artificial Intelligence of Things) that structures and formalizes the main features of such systems. The ultimate goal is to provide a framework that guides the selection of the most suitable XAI solution for a given intelligent IoT application. This conceptual model is validated by classifying up to 100 approaches obtained from a systematic review of the state-of-the-art and formalized in an ontology to guide XAIoT system design.
- Open Access
- 30 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Explainable Deep Learning for Multi-Label ECG Classification: Clinical Validation and Analysis of Signal Transformations
Deep learning models for automated electrocardiogram (ECG) interpretation have achieved high diagnostic accuracy, yet clinical adoption remains limited due to a lack of explainability and frequent reliance on private datasets and closed-source implementations. We present a clinically validated explainable AI (XAI) framework for multi-label ECG classification that shifts the focus from raw performance toward explainability, reproducibility, and clinical usability. Our framework is developed exclusively using publicly available resources—the PTB-XL database (21,799 ECGs across five diagnostic categories), an open-source model (Ribeiro et al. [14]), and open-source libraries. The model achieves an F1-score of 0.738 while providing gradient-based saliency maps evaluated by practicing cardiologists. We further compare time-domain and time–frequency representations (STFT and CWT). Direct time-series analysis achieves superior performance compared to transformed inputs with this model, while also requiring substantially lower computational cost. A key finding from a two-round validation process with clinicians is that explanation usefulness depends not only on the attribution method but also on presentation format: cardiologists could not reliably interpret saliency overlays until they were rendered on standard ECG grid paper used in daily clinical practice. Our contributions are: (1) a fully reproducible XAI pipeline with documented iterative clinical validation, (2) evidence that visualization format is integral to explanation quality in ECG analysis, (3) a GAM-compliant extension of the open-source ecg_plot library, and (4) a controlled analysis of signal transformation trade-offs.
- 30 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessExplainable Deep Learning for Multi-Label ECG Classification: Clinical Validation and Analysis of Signal Transformations
Deep learning models for automated electrocardiogram (ECG) interpretation have achieved high diagnostic accuracy, yet clinical adoption remains limited due to a lack of explainability and frequent reliance on private datasets and closed-source implementations. We present a clinically validated explainable AI (XAI) framework for multi-label ECG classification that shifts the focus from raw performance toward explainability, reproducibility, and clinical usability. Our framework is developed exclusively using publicly available resources—the PTB-XL database (21,799 ECGs across five diagnostic categories), an open-source model (Ribeiro et al. [14]), and open-source libraries. The model achieves an F1-score of 0.738 while providing gradient-based saliency maps evaluated by practicing cardiologists. We further compare time-domain and time–frequency representations (STFT and CWT). Direct time-series analysis achieves superior performance compared to transformed inputs with this model, while also requiring substantially lower computational cost. A key finding from a two-round validation process with clinicians is that explanation usefulness depends not only on the attribution method but also on presentation format: cardiologists could not reliably interpret saliency overlays until they were rendered on standard ECG grid paper used in daily clinical practice. Our contributions are: (1) a fully reproducible XAI pipeline with documented iterative clinical validation, (2) evidence that visualization format is integral to explanation quality in ECG analysis, (3) a GAM-compliant extension of the open-source ecg_plot library, and (4) a controlled analysis of signal transformation trade-offs.
- Open Access
- 1 January 2026
- Published by Walter de Gruyter GmbH in Journal of Intelligent Systems
- Vol. 35 (1)
Unveiling the potential of semantically-backed behavior trees for retrieving explanation experiences: a comparative study
Due to endeavors in the field of eXplainable Artificial Intelligence (XAI), designers of AI systems face an ever-growing plethora of methods for making the predictions and decisions of their systems more human interpretable. This paper unveils results from the iSee project, which uses Case-Based Reasoning to assist designers in selecting among these methods. In iSee, each case within the case base encapsulates a user’s experience employing one or more methods to fulfill an explanation requirement. Specifically, we refer to individual explanation methods as explainers , and the composition of explainers (which may be needed to satisfy more complex explanation needs) as explanation strategies . In this paper, we elucidate our utilization of Behavior Trees (BTs) for representing the workflow inherent in each explanation strategy, where the vocabulary of BT nodes is defined by the iSeeOnto ontology. AI system designers delineate their explanation needs, enabling the iSee platform to retrieve similar cases and adapt the BTs to cater to the user’s unique needs. This paper further conducts a comparative analysis of different ways of measuring similarity between iSee BTs. We propose several combinations of measures of edit distance with ontological distance. By combining these two, our similarity measures capture not just the structure of the trees but also the semantics of the nodes within the trees. The results should also be of interest to researchers who want to measure the similarity of trees when an ontology defines the semantics of the nodes. The evaluation of the approach was conducted using two experimental protocols (leave-one-out and leave-one-in) over a synthetic iSee case base, where the proposed similarity measures for BT adaptation were assessed. The results support our hypothesis, showing that adapting BTs leveraging not only structural but also semantic knowledge improves the retrieval and reuse of explanation strategies. These findings suggest that incorporating semantic-aware similarity into workflow adaptation can enhance the effectiveness of explanation design in XAI systems.
- 1 January 2026
- Published by Walter de Gruyter GmbH in Journal of Intelligent Systems
- Vol. 35 (1)
Open AccessUnveiling the potential of semantically-backed behavior trees for retrieving explanation experiences: a comparative study
Due to endeavors in the field of eXplainable Artificial Intelligence (XAI), designers of AI systems face an ever-growing plethora of methods for making the predictions and decisions of their systems more human interpretable. This paper unveils results from the iSee project, which uses Case-Based Reasoning to assist designers in selecting among these methods. In iSee, each case within the case base encapsulates a user’s experience employing one or more methods to fulfill an explanation requirement. Specifically, we refer to individual explanation methods as explainers , and the composition of explainers (which may be needed to satisfy more complex explanation needs) as explanation strategies . In this paper, we elucidate our utilization of Behavior Trees (BTs) for representing the workflow inherent in each explanation strategy, where the vocabulary of BT nodes is defined by the iSeeOnto ontology. AI system designers delineate their explanation needs, enabling the iSee platform to retrieve similar cases and adapt the BTs to cater to the user’s unique needs. This paper further conducts a comparative analysis of different ways of measuring similarity between iSee BTs. We propose several combinations of measures of edit distance with ontological distance. By combining these two, our similarity measures capture not just the structure of the trees but also the semantics of the nodes within the trees. The results should also be of interest to researchers who want to measure the similarity of trees when an ontology defines the semantics of the nodes. The evaluation of the approach was conducted using two experimental protocols (leave-one-out and leave-one-in) over a synthetic iSee case base, where the proposed similarity measures for BT adaptation were assessed. The results support our hypothesis, showing that adapting BTs leveraging not only structural but also semantic knowledge improves the retrieval and reuse of explanation strategies. These findings suggest that incorporating semantic-aware similarity into workflow adaptation can enhance the effectiveness of explanation design in XAI systems.
- Open Access
- 8 December 2025
- Published by IBERAMIA: Sociedad Iberoamericana de Inteligencia Artificial in INTELIGENCIA ARTIFICIAL
- Vol. 29 (77) , 13-27
From the algorithm to the clinical interpretation of childbirth anxiety: analysis and explainability of obstetric predictive models based on psychological indicators
Anxiety during pregnancy constitutes a relevant factor that can significantly influence labor development. This study presents a novel approach based on explainable artificial intelligence to predict both the type and duration of labor using psychological indicators of anxiety prior to delivery. Employing data from 235 full-term pregnant women from two Spanish hospitals, we developed a multilayer perceptron model to classify eutocic and dystocic deliveries, achieving a capacity to identify 88\% of dystocic deliveries. Additionally, we implemented a regression model that predicts labor time with a mean error of 2 hours, correctly predicting 86% of cases with an error margin of less than 3 hours. The application of explainability techniques to the developed models allows for understanding the specific influence of each anxiety factor on labor development. These results demonstrate the potential of AI models to improve obstetric care and optimize healthcare resource allocation.
- 8 December 2025
- Published by IBERAMIA: Sociedad Iberoamericana de Inteligencia Artificial in INTELIGENCIA ARTIFICIAL
- Vol. 29 (77) , 13-27
Open AccessFrom the algorithm to the clinical interpretation of childbirth anxiety: analysis and explainability of obstetric predictive models based on psychological indicators
Anxiety during pregnancy constitutes a relevant factor that can significantly influence labor development. This study presents a novel approach based on explainable artificial intelligence to predict both the type and duration of labor using psychological indicators of anxiety prior to delivery. Employing data from 235 full-term pregnant women from two Spanish hospitals, we developed a multilayer perceptron model to classify eutocic and dystocic deliveries, achieving a capacity to identify 88\% of dystocic deliveries. Additionally, we implemented a regression model that predicts labor time with a mean error of 2 hours, correctly predicting 86% of cases with an error margin of less than 3 hours. The application of explainability techniques to the developed models allows for understanding the specific influence of each anxiety factor on labor development. These results demonstrate the potential of AI models to improve obstetric care and optimize healthcare resource allocation.
- 7 November 2025
- Published by Springer Nature
- p. 269-280
A Practical Framework for Auditing Fairness in Medical AI
- 7 November 2025
- Published by Springer Nature
- p. 269-280
A Practical Framework for Auditing Fairness in Medical AI
- Open Access
- 25 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 71-79
Binary Mexican Axolotl Optimization (BMAO): A Bio-Inspired Algorithm for Enhancing Machine Learning Performance through Feature Selection
- 25 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 71-79
Open AccessBinary Mexican Axolotl Optimization (BMAO): A Bio-Inspired Algorithm for Enhancing Machine Learning Performance through Feature Selection