Conferences
Contribution |
Link |
Abstract |
| Patricia Camacho Magriñán, Diego Sales Lérida, Antonio León Jiménez, Alejandro Lara Doña, Eva Vázquez Gandullo, Alfonso Marín Andreu, Aurelio Arnedillo Muñoz, Daniel Sánchez Morillo. A Longitudinal Multivariate Assessment of Acoustic and Motor Biomarkers in COPD Telehealth. In The International Conference on Artificial Intelligence in Health (aiHealth 2026), Lesvos, Grecia, 2026 | Link | Chronic Obstructive Pulmonary Disease (COPD) remains a leading cause of global morbidity and mortality. Disease-related exacerbations (AECOPD) impose a substantial socioeconomic burden and trigger irreversible declines in lung function. While COPD telemonitoring and artificial intelligence has emerged as a pivotal tool for self-management and early exacerbation detection, its clinical efficacy in real-world settings is inherently contingent upon patient adherence. A critical, yet underexplored, challenge in telehealth is the “adherence dichotomy”: the tension between low-friction passive data collection and the high cognitive and physical burden of active self-reporting. The primary objective of this study was to quantify technical and clinical compliance within a multimodal daily monitoring protocol for COPD. We sought to characterize longitudinal adherence patterns and evaluate how task complexity and physiological strain influence patient retention. This analysis could support the optimization of the user experience while ensuring the high-granularity data streams required for robust predictive modeling. |
| Daniel Sánchez Morillo, Patricia Camacho Magriñán, Diego Sales Lérida, Alejandro Lara Doña, Eva Vázquez Gandullo, Alfonso Marín Andreu, Aurelio Arnedillo Muñoz, Antonio León Jiménez. Multimodal COPD Monitoring Ecosystems: Adherence Patterns as a Barrier to AI-Driven Exacerbation Prediction. In The International Conference on Artificial Intelligence in Health (aiHealth 2026), Lesvos, Grecia, 2026. | Continuous patient monitoring using digital biomarkers offers a significant opportunity for preventing clinical deterioration and improving preventive decision-making [1]. In chronic diseases such as COPD, the ability to identify the prodromal phase (the period preceding an exacerbation) is important for risk reduction; however, this is challenged by high inter-patient variability. The primary objective of this study is to evaluate the predictive capacity of various digital variables, acoustic and fine motor in nature, to differentiate between stable and prodromal states in patients with COPD. We hypothesize that subtle alterations in phonation, speech rhythm, and fine psychomotor function act as early, objective indicators of an imminent exacerbation. | |
| P. C. Magriñán, C. Jácome and D. S. Morillo, “Usability Evaluation of a Clinical Dashboard for the Telemonitoring of Patients with COPD,” 2026 6th International Conference on Electrical, Computer and Energy Technologies (ICECET), Rome, Italy, 2026, pp. 1-6, | DOI | This study evaluates the usability of a clinical dashboard developed under the CICERONE Project for the remote monitoring of patients with Chronic Obstructive Pulmonary Disease (COPD). The platform, built on Power BI, integrates multidimensional data streams—including physiological, environmental, physical activity, sleep, and Patient-Reported Outcomes (PROs). Using a mixed-methods approach with a cohort of six medical professionals, the system yielded a System Usability Scale (SUS) score of 77.92, denoting “good” usability and validating the underlying information architecture, despite inter-rater variability. Our analysis reveals a critical divergence between perceived clinical utility (mean ≈4.3) and interaction efficiency (mean ≈2.8). This suggests that clinicians may tolerate interaction friction in exchange for high functional value. However, a significant deficit was identified in PRO visualization clarity (mean ≈2.7), indicating the need for redesign to optimize data interpretability and clinical decision-making. These findings provide a framework for the iterative transition of the dashboard into a dedicated web application. |
| Camacho-Magriñán, P., Sales-Lérida, D., Priego-Torres, B., Sánchez-Morillo, D. (2026). Monitoring and Predicting Indoor Air Quality: An IoT and Machine Learning-Based Approach. In: Ullah, A., Anwar, S. (eds). Proceedings of International Conference on Information Technology and Applications. ICITA 2025. Lecture Notes in Networks and Systems, vol 1889. Springer, Cham. | DOI | Indoor Air Quality (IAQ) is crucial for human health, especially in homes where people spend most of their time. Common pollutants like VOCs (Volatile Organic Compounds) and particulate matter (PM), often produced by activities like cooking, can have a negative health impact, while proper ventilation is essential to reduce these risks. Growing interest in IAQ has led to the development of low-cost IoT-based monitoring systems using affordable sensors to track key parameters such as PM, CO, VOCs, temperature, and humidity. While these systems are effective for monitoring, limitations remain in accurately quantifying individual exposure. Furthermore, the ability to predict air quality is crucial to enable preventive actions. However, existing studies often lack comprehensive temporal analyses, focus on limited pollutants, or do not assess compliance with World Health Organization (WHO) guidelines, highlighting the need for more robust and integrated approaches. This study presents an IoT-based system for monitoring IAQ, collecting data on various pollutants in home kitchens. Data was gathered over seven days in three kitchens. The research includes time-based analysis, multivariate predictive models, and comparison with WHO standards. Strengths of this study include the simultaneous monitoring of multiple pollutants and the use of predictive models to support real-time decision-making, while limitations involve the relatively short monitoring period and potential variability between households. This comprehensive approach aims to enhance understanding of IAQ in kitchens, identify gaps in current knowledge, and facilitate proactive decision-making for the health and safety of occupants. |
| Patricia Camacho Magriñán, Daniel Sánchez Morillo, Regla Moreno Mellado, Eva Vázquez Gandullo, Alfonso Marín Andreu, Antonio León Jiménez. Usability Study of the CICERONE App for Telemonitoring COPD Patients. In 2025, IARIA Annual Congress on Frontiers in Science, Technology, Services, and Applications (IARIA CONGRESS 2025), July 06-10, 2025, Venice, Italy. | Link | Telemedicine is a promising tool for the management of Chronic Obstructive Pulmonary Disease (COPD), but its implementation faces challenges related to, among other issues, patient acceptance and usability. In this context, the effectiveness of telemonitoring systems depends not only on their accuracy in predicting exacerbations but also on their ability to integrate intuitively and efficiently into users’ daily lives. The CICERONE project has developed a multimodal home telemonitoring system for COPD patients that collects data on symptoms, environmental parameters, lifestyle, and biomedical information. This work presents a study focused on the system’s usability, evaluated in three stages: a) initial testing of a mock-up with patients; b) functional trials with volunteers; and c) final testing with patients participating in the project. The development followed an iterative approach based on user feedback and evaluations using the System Usability Scale (SUS). The results highlight how iteration and user-centered design have improved the patient experience and optimized the system’s functionality. This study underscores the importance of usability in the design of telemonitoring tools to ensure their adoption and effectiveness in real clinical settings, promoting a more personalized and proactive approach to COPD management. |
| Patricia Camacho Magriñán, Daniel Sánchez Morillo, Eva Vazquez-Gandullo, Alfonso Marín-Andreu, Antonio León-Jimenez, AI, smart sensors and new predictors for COPD management. Natural Science Baltic Conference, Poland, 2025. | Link | Chronic Obstructive Pulmonary Disease (COPD) is a major cause of morbidity and mortality worldwide. Telemonitoring plays a significant role in the management of COPD since it enables healthcare providers to detect early signs of COPD exacerbations, allowing for prompt intervention, and potentially preventing hospitalizations. While existing telemonitoring approaches have shown promise in reducing hospitalization rates, further studies are needed to evaluate the impact of telemonitoring on quality of life. Through the CICERONE project, we introduce a home telemonitoring system that gathers multimodal data. This study presents the first developments, including a mobile application, an environmental air quality monitoring system, and an electronic patient record that brings all heterogeneous data in one place. This app has been developed using a user-centered design methodology with a group of 31 participants and empowers patients through surveys, games, voice and cough sound recording, and integration with vital sensors (peak flow meter, smart wristband, and a scale). Collected data is transmitted to a centralized electronic record, enabling continuous patient monitoring. Currently, the system is gathering data in an observational study with a group of COPD patients. Machine learning models will be trained and validated to enhance the early detection of imminent COPD exacerbations. We aim to contribute to the advancement of COPD management by developing tools that foster proactive and individualized care. |
| Patricia Camacho Magriñán, Daniel Sánchez Morillo. Datos multimodales e Inteligencia Artificial en Salud Respiratoria: Predicción de Exacerbaciones de la EPOC. En la Jornada de Defensa de Planes de Investigación, Programa de Doctorado en Fabricación, Materiales e Ingeniería Ambiental de la Universidad de Cádiz. 2024. | DOI | |
| Patricia Camacho Magriñán, Diego Sales Lérida, Antonio León Jiménez, Daniel Sánchez Morillo. Impacto de la Calidad del Aire Interior en Viviendas: Estudio Piloto con Sensores IoT de Bajo Coste, En XLIII Congreso Anual de la Sociedad Española de Ingeniería Biomédica. CASEIB 2025: Libro de actas, 2025. | Link | Este trabajo presenta un estudio piloto sobre la calidad del aire interior en viviendas mediante el desarrollo y validación de Cicero-ne AirLink, un sistema IoT de bajo coste que integra sensores de PM, COVs y CO2 con conectividad NB-IoT. La monitorización se realizó en nueve viviendas de la provincia de Cádiz durante siete días, con dispositivos instalados en cocina, salón y dormitorio. Los resultados evidencian que la calidad del aire interior es altamente dinámica y dependiente de la actividad humana y la ventilación, identificándose la cocina como punto crítico de emisión de partículas y COVs durante el cocinado, y los dormitorios como espacios de acumulación significativa de CO2, con picos cercanos a 3000 ppm. El análisis estadístico mostró diferencias significativas entre estancias y franjas horarias, mientras que los modelos predictivos basados en Random Forest alcanzaron una elevada precisión en la estimación de contaminantes, lo que demuestra su utilidad para anticipar episodios de exposición y habilitar estrategias de ventilación inteligente. Estos hallazgos confirman la viabilidad técnica y económica de los sensores de bajo coste combinados con analítica avanzada como herramienta de apoyo a la salud pública y al diseño de entornos residenciales más saludables. |
| Patricia Camacho Magriñán, Diego Sales Lérida, Regla Moreno Mellado, Alejandro Lara Doña, Daniel Sánchez Morillo. Sistema de telemonitorización de la calidad de aire interior para pacientes con EPOC, En XLII Congreso Anual de la Sociedad Española de Ingeniería Biomédica. CASEIB 2024: Libro de actas, 2024. | Link | La Enfermedad Pulmonar Obstructiva Crónica (EPOC) es una causa principal de mortalidad global, con exacerbaciones que suponen una significativa carga socioeconómica. La calidad del aire, tanto interior como exterior, influye significativamente en la salud respiratoria. Este estudio revisa la evidencia científica que relaciona la calidad del aire interior y la EPOC, subrayando la importancia de la monitorización en tiempo real mediante tecnologías avanzadas e inteligencia artificial (IA). Para ello, se ha realizado una revisión sistemática de la literatura, identificando estudios recientes que emplean IA para predecir exacerbaciones de EPOC a partir de datos ambientales y clínicos. Los resultados muestran que contaminantes como el material particulado (PM), los compuestos orgánicos volátiles (COVs) y el dióxido de carbono (CO2) representan un factor de riesgo para le EPOC, y que la integración de técnicas de IA tiene potencial para anticipar episodios de exacerbación, habilitando la intervención temprana y mejorando la calidad de vida de los pacientes. El conocimiento adquirido a partir de los estudios analizados ha sido empleado para diseñar y fabricar un sistema sensorizado de bajo coste equipado con sensores de CO2, PM y COVs. |
| Diego Sales Lérida, Patricia Camacho Magriñán, Regla Moreno Mellado, Alejandro Lara Doña, Blanca Priego Torres, Aurelio Arnedillo Muñoz, Antonio León Jiménez, Daniel Sánchez Morillo,. Inteligencia artificial, sensores inteligentes y nuevos predictores fisiológicos y medioambientales para una mejor gestión de la EPOC , En XLII Congreso Anual de la Sociedad Española de Ingeniería Biomédica. CASEIB 2024: Libro de actas, 2024. | Link | La telemedicina ha demostrado ser una herramienta prometedora para prevenir ingresos hospitalarios por exacerbaciones de la enfermedad pulmonar obstructiva crónica (EPOC). Sin embargo, los resultados de los ensayos de telemonitorización con pacientes con EPOC revelan una falta de especificidad en los modelos predictivos desarrollados, que generan frecuentes alertas clínicamente innecesarias, con el consiguiente consumo excesivo de recursos sanitarios o con niveles evitables de estrés y ansiedad para los pacientes y cuidadores. Existe por tanto la necesidad de encontrar nuevos predictores de exacerbaciones y de desarrollar modelos predictivos. El proyecto CICERONE aborda estos objetivos mediante la búsqueda de nuevos predictores fisiológicos y ambientales de exacerbaciones de EPOC que habiliten el desarrollo de modelos basados en inteligencia artificial fiable. En este estudio, se ha desarrollado un sistema de telemonitorización domiciliaria multimodal para pacientes de EPOC, que registra síntomas, datos ambientales y de estilo de vida, e información biomédica. Se presenta en este trabajo el sistema desarrollado, junto a aspectos metodológicos y de usabilidad del desarrollo, que será empleado en un estudio de campo con un grupo de pacientes de EPOC. El objetivo final es el de desplazar el manejo de la EPOC de un enfoque genérico a uno personalizado, a través de la investigación de parámetros fisiológicos para la predicción de exacerbaciones de la EPOC y la generación de modelos predictivos que ofrezcan un compromiso entre la comprensión clínica y la bondad del ajuste estadístico. |
Grant PID2021-126810OB-I00 funded by:
