Lucian Georghe GRUIONU
Lucian Gheorghe Gruionu, Professor, PhD, Eng., habil., is a senior academic and researcher in mechanical and biomedical engineering at the Faculty of Mechanics, University of Craiova, Romania. His research activity over more than two decades has been focused on medical robotics, image-guided interventions, biomechanics, and the development of advanced medical instruments and intelligent diagnostic systems.
Professor Gruionu has extensive international research experience, having worked as a researcher and collaborator at prestigious institutions in the United States, including Johns Hopkins University, Georgetown University, and Indiana University School of Medicine. His work has been carried out in close collaboration with clinicians, addressing real clinical needs in minimally invasive procedures, particularly in bronchoscopy, endoscopy, interventional radiology, and oncologic diagnostics.
He has served as principal investigator and project director for numerous national and international research grants, including large collaborative projects funded through European and Norwegian mechanisms, focusing on artificial intelligence, robotic navigation, and innovative medical devices. His research has led to the design and validation of novel robotic platforms for flexible instrument navigation, as well as AI-based systems for radiation-free guidance in complex anatomical environments.
Professor Gruionu is the author and co-author of over one hundred peer-reviewed journal articles, conference papers, and book chapters, and he is an inventor of patented medical devices in the field of medical robotics and image-guided interventions. In parallel with his academic activity, he is co-founder and CEO of an R&D-oriented company, actively involved in the translation of research results toward clinical and industrial applications.
His current research interests lie at the intersection of robotics, sensing technologies, and artificial intelligence, with the objective of developing safer, smarter, and clinically relevant systems that enhance physician capabilities and improve patient outcomes in minimally invasive medicine.
The title of the presentation: From Robotic Catheter Control to AI Shape Sensing in Lung Bronchoscopy
Abstract
Early diagnosis of peripheral lung cancer remains a major clinical challenge due to the limited reach of conventional bronchoscopes and the heavy reliance on fluoroscopy or electromagnetic tracking, exposing both patients and clinicians to ionizing radiation. This talk presents a unified robotic and AI-driven framework for safe, precise, and radiation-free navigation of flexible instruments inside the lung airways.
The presentation integrates two complementary technological directions developed and validated by our research group:
- a compact robotic platform (RoboCath) for precise manipulation of long, flexible catheters during bronchoscopy, designed to seamlessly integrate with standard bronchoscopes. The robot enables controlled translation and rotation of instruments beyond the bronchoscope tip, significantly improving access to peripheral lung regions while reducing procedural complexity and X-ray dependency. Extensive feasibility testing in anatomically accurate lung phantoms demonstrated reliable navigation across multiple bronchial generations with a small operating-room footprint and sterilization-compatible design
- a novel shape-sensing navigation system based on Fiber Bragg Grating (FBG) technology and artificial intelligence (AIrShape), capable of tracking the entire catheter shape in real time. By matching the measured catheter geometry to precomputed airway centerlines using a multi-view convolutional neural network, the system identifies the active airway without fluoroscopy or electromagnetic tracking. Experimental validation showed a mean airway identification accuracy of approximately 91%, enabling safe navigation toward peripheral lung lesions beyond the reach of conventional imaging guidance
Together, these results demonstrate how robotic actuation, optical shape sensing, and AI-based anatomical reasoning can be combined into next-generation bronchoscopy platforms. The proposed technologies pave the way for radiation-free, cost-effective, and scalable robotic systems for early lung cancer diagnosis, with strong potential for future preclinical and clinical translation.
Professor Gruionu has extensive international research experience, having worked as a researcher and collaborator at prestigious institutions in the United States, including Johns Hopkins University, Georgetown University, and Indiana University School of Medicine. His work has been carried out in close collaboration with clinicians, addressing real clinical needs in minimally invasive procedures, particularly in bronchoscopy, endoscopy, interventional radiology, and oncologic diagnostics.
He has served as principal investigator and project director for numerous national and international research grants, including large collaborative projects funded through European and Norwegian mechanisms, focusing on artificial intelligence, robotic navigation, and innovative medical devices. His research has led to the design and validation of novel robotic platforms for flexible instrument navigation, as well as AI-based systems for radiation-free guidance in complex anatomical environments.
Professor Gruionu is the author and co-author of over one hundred peer-reviewed journal articles, conference papers, and book chapters, and he is an inventor of patented medical devices in the field of medical robotics and image-guided interventions. In parallel with his academic activity, he is co-founder and CEO of an R&D-oriented company, actively involved in the translation of research results toward clinical and industrial applications.
His current research interests lie at the intersection of robotics, sensing technologies, and artificial intelligence, with the objective of developing safer, smarter, and clinically relevant systems that enhance physician capabilities and improve patient outcomes in minimally invasive medicine.
The title of the presentation: From Robotic Catheter Control to AI Shape Sensing in Lung Bronchoscopy
Abstract
Early diagnosis of peripheral lung cancer remains a major clinical challenge due to the limited reach of conventional bronchoscopes and the heavy reliance on fluoroscopy or electromagnetic tracking, exposing both patients and clinicians to ionizing radiation. This talk presents a unified robotic and AI-driven framework for safe, precise, and radiation-free navigation of flexible instruments inside the lung airways.
The presentation integrates two complementary technological directions developed and validated by our research group:
- a compact robotic platform (RoboCath) for precise manipulation of long, flexible catheters during bronchoscopy, designed to seamlessly integrate with standard bronchoscopes. The robot enables controlled translation and rotation of instruments beyond the bronchoscope tip, significantly improving access to peripheral lung regions while reducing procedural complexity and X-ray dependency. Extensive feasibility testing in anatomically accurate lung phantoms demonstrated reliable navigation across multiple bronchial generations with a small operating-room footprint and sterilization-compatible design
- a novel shape-sensing navigation system based on Fiber Bragg Grating (FBG) technology and artificial intelligence (AIrShape), capable of tracking the entire catheter shape in real time. By matching the measured catheter geometry to precomputed airway centerlines using a multi-view convolutional neural network, the system identifies the active airway without fluoroscopy or electromagnetic tracking. Experimental validation showed a mean airway identification accuracy of approximately 91%, enabling safe navigation toward peripheral lung lesions beyond the reach of conventional imaging guidance
Together, these results demonstrate how robotic actuation, optical shape sensing, and AI-based anatomical reasoning can be combined into next-generation bronchoscopy platforms. The proposed technologies pave the way for radiation-free, cost-effective, and scalable robotic systems for early lung cancer diagnosis, with strong potential for future preclinical and clinical translation.
HAROON
Haroon is a Lecturer at Chalmers University. Previously, Haroon was a visiting research scholar at the Department of Computer Science and Engineering, Southern University of Technology, China, where he worked on AI regulation. He identified challenges to effective regulation of AI systems and proposed conceptual, methodological, and practical solutions to address these challenges. Before that, he had a postdoctoral position at Umeå University, Sweden. He conducted research in software security with a particular focus on fuzzer evaluation. He discovered serious flaws in state-of-the-art fuzzer evaluation benchmarks and proposed mitigations. He completed his doctoral degree at the School of Computer Science, Guangzhou University, China. The focus of his PhD research was on data over-collection in Android smartphones.
His research interests include identifying and solving privacy, security, and trust issues of emerging technologies in the rapidly changing privacy and security threat landscapes. He is also interested in identifying and solving problems arising from placing the solutions based on these technologies in operational settings. He has published over twenty articles and conference papers at reputed venues, including IEEE TDSC, IEEE TCSS, IEEE IoT Journal, IEEE TVT, Information Sciences, Computer Standards and Interfaces, and Neurocomputing.
His research interests include identifying and solving privacy, security, and trust issues of emerging technologies in the rapidly changing privacy and security threat landscapes. He is also interested in identifying and solving problems arising from placing the solutions based on these technologies in operational settings. He has published over twenty articles and conference papers at reputed venues, including IEEE TDSC, IEEE TCSS, IEEE IoT Journal, IEEE TVT, Information Sciences, Computer Standards and Interfaces, and Neurocomputing.
The title of the presentation: On Regulating High-Risk AI Systems
Abstract
Regulating high-risk artificial intelligence (AI) systems is an urgent issue, yet technical infrastructure for their effective regulation remains scarce. In this paper, we address this gap
by identifying key challenges in developing technical frameworks for AI systems’ regulation and proposing conceptual, methodological, and practical solutions to address these challenges. In this regard, we introduce the concept of AI’s operational qualification and propose the temporal self-replacement test, akin to certification tests for human operators, to examine the AI’s operational qualification. We propose measuring AI’s operational qualification across its operational properties critical for its regulatory fitness and introduce the operational qualification score as a pragmatic measure of AI’s regulatory fitness. In addition, we design and develop a Secure Framework for AI Regulation (SFAIR), a tool for automatic, recurrent, and secure examination of an AI’s operational qualification and attestation of its regulatory fitness, leveraging the proposed test and measure. We validate the efficacy of the temporal self-replacement test and the practical utility of SFAIR by demonstrating its capability to support regulatory authorities in automated, recurrent, and secure AI qualification examination and attestation of its regulatory fitness using an open-source, high-risk AI system. Finally, we make the source code of SFAIR publicly available.
Abstract
Regulating high-risk artificial intelligence (AI) systems is an urgent issue, yet technical infrastructure for their effective regulation remains scarce. In this paper, we address this gap
by identifying key challenges in developing technical frameworks for AI systems’ regulation and proposing conceptual, methodological, and practical solutions to address these challenges. In this regard, we introduce the concept of AI’s operational qualification and propose the temporal self-replacement test, akin to certification tests for human operators, to examine the AI’s operational qualification. We propose measuring AI’s operational qualification across its operational properties critical for its regulatory fitness and introduce the operational qualification score as a pragmatic measure of AI’s regulatory fitness. In addition, we design and develop a Secure Framework for AI Regulation (SFAIR), a tool for automatic, recurrent, and secure examination of an AI’s operational qualification and attestation of its regulatory fitness, leveraging the proposed test and measure. We validate the efficacy of the temporal self-replacement test and the practical utility of SFAIR by demonstrating its capability to support regulatory authorities in automated, recurrent, and secure AI qualification examination and attestation of its regulatory fitness using an open-source, high-risk AI system. Finally, we make the source code of SFAIR publicly available.
OCTAVIAN ADRIAN POSTOLACHE
Dr. Octavian Adrian Postolache (M’99, SM’06) graduated in Electrical Engineering at the Gh. Asachi Technical University of Iasi, Romania, in 1992 and he received the PhD degree in 1999 from the same university, and university habilitation in 2016 from Instituto Superior Tecnico, Universidade de Lisboa, Portugal. He joined Instituto Universitario de Lisboa/ ISCTE-IUL Lisbon where he is currently Associate Professor . His fields of interests are smart sensors for biomedical and environmental applications, pervasive sensing and computing, wireless sensor networks, signal processing with application in biomedical and telecommunications, computational intelligence with application in automated measurement systems. Dr. Postolache is author and co-author of 9 patents, 10 books, 22 book chapters, 107 papers in international journals with peer review, more than 295 papers in proceedings of international conferences with peer review. He is IEEE Senior Member I&M Society, Distinguished Lecturer of IEEE IMS 2017-2020, chair of IEEE I&MSTC-13 Wireless and Telecommunications in Measurements, member of IEEE I&M TC-17, IEEE I&M TC-18, IEEE I&MS TC-25, IEEE EMBS Portugal Chapter and chair of IEEE IMS Portugal Chapter. He is Associate Editor of IEEE Sensors Journal, and IEEE Transaction on Instrumentation and Measurements, he was general chair of an important number of IEEE conferences. He received IEEE Sensors Journal best reviewer and the best associate editor in 2011, 2013 and 2017, and other awards related to his research activity in the field of smart sensing.
The title of the presentation: Digital Transformation in Healthcare: Smart Physical Therapy
Abstract
The convergence of healthcare, instrumentation and measurement technologies will transform healthcare as we know it, improving quality of healthcare services, reducing inefficiencies, curbing costs and improving quality of life. Smart sensors, wearable devices, Internet of Things (IoT) platforms, and big data offer new and exciting possibilities for more robust, reliable, flexible and low-cost healthcare systems and patient care strategies. The data coming from the rehabilitation process is useful to create AI personalized models associated with physical rehabilitation plans optimization, patient outcome prediction, clinics resource optimization.
This tutorial highlights the development of rehabilitation solutions based on smart sensors virtual reality and serious games. As part of these interactive environments, 3D image sensors will be introduced for natural user interaction with rehabilitation scenarios and remote sensing of the user movements, along with thermal cameras for remote evaluation of muscle activity. Additionally, non-invasive monitoring technologies for tracking patients' posture, balance, and gait during the rehabilitation process will be presented. Developed prototypes, such as smart walkers and force platforms, will be discussed, providing quantitative insights related to physical rehabilitation outcomes.
The tutorial will also address challenges related to signal processing, data storage, representation, and analysis, including the formulation of specific metrics for assessing patient progress throughout the rehabilitation process. Elements regarding AI modeling, and AI implementation are considered taking into account that AI may provide to the clinical specialists knowledges about performance metrics after every training session, helping them better understand the motor limitations of the patient and the latest improvements.
The title of the presentation: Digital Transformation in Healthcare: Smart Physical Therapy
Abstract
The convergence of healthcare, instrumentation and measurement technologies will transform healthcare as we know it, improving quality of healthcare services, reducing inefficiencies, curbing costs and improving quality of life. Smart sensors, wearable devices, Internet of Things (IoT) platforms, and big data offer new and exciting possibilities for more robust, reliable, flexible and low-cost healthcare systems and patient care strategies. The data coming from the rehabilitation process is useful to create AI personalized models associated with physical rehabilitation plans optimization, patient outcome prediction, clinics resource optimization.
This tutorial highlights the development of rehabilitation solutions based on smart sensors virtual reality and serious games. As part of these interactive environments, 3D image sensors will be introduced for natural user interaction with rehabilitation scenarios and remote sensing of the user movements, along with thermal cameras for remote evaluation of muscle activity. Additionally, non-invasive monitoring technologies for tracking patients' posture, balance, and gait during the rehabilitation process will be presented. Developed prototypes, such as smart walkers and force platforms, will be discussed, providing quantitative insights related to physical rehabilitation outcomes.
The tutorial will also address challenges related to signal processing, data storage, representation, and analysis, including the formulation of specific metrics for assessing patient progress throughout the rehabilitation process. Elements regarding AI modeling, and AI implementation are considered taking into account that AI may provide to the clinical specialists knowledges about performance metrics after every training session, helping them better understand the motor limitations of the patient and the latest improvements.
OANA GEMAN
Dr. Oana Geman ((IEEE SM ‘18) is Medical Bioengineer and PhD in Electronics and Telecommunication and a post-doctoral researcher in Computer Science. She is currently a Senior Teaching Fellow (Professor) at Division of Data Science and Artificial Intelligence, Computer Science and Engineering, Chalmers University of Technology and University of Gothenburg, Sweden and an Associate Professor at University of Suceava, Romania. She obtained Habilitation in Electronics and Telecommunication Field in 2018. Within the past five years she published 10 books, has published over 100 articles in ISI Web of Science journals, with FI over 40), and her various works have been cited over 4000 times. She served as Chair of many Internationals Conferences or Organizer, Session Chair and member in Program or Technical Committees and a Senior Member IEEE. She has been a director or a member in 10 national and international grants. She was considered three years in the row by Stanford - Elsevier, in the top 2% of the highest cited scientists. Her current research interests include: non-invasive measurements of biomedical signals, wireless sensors, signal processing, and processing information by way of Artificial Intelligence such as nonlinear dynamics analysis, stochastic networks and neuro-fuzzy methods, classification and prediction, Data-Mining, Deep Learning, Intelligent Systems etc. She is a reviewer and editor of many top journals, including IEEE Transactions, IEEE ACCESS, AIHC Journal, IOT Journal, Sensors, Springer Nature Journals etc.
The title of the presentation: Machine Learning and Deep Learning in Brain Research
Abstract
The integration of neuroscience with artificial intelligence is transforming the way we study and understand the human brain. Machine Learning (ML) and Deep Learning (DL) offer powerful methods for making sense of the vast and complex data generated by modern brain research, from neuroimaging and electrophysiological recordings to behavioral measurements. These technologies enable more accurate diagnoses, earlier detection of neurological disorders, and deeper insights into how the brain functions in health and disease.
This lecture explores how ML and DL are applied in key areas of brain research, including neuroimaging analysis, brain–computer interfaces, and cognitive modeling. It introduces practical approaches for working with EEG, fMRI, MEG, and multimodal datasets, focusing on feature extraction, pattern discovery, and end-to-end deep learning models. Neural network architectures such as convolutional and recurrent networks are discussed in the context of brain state classification, disease prediction (ex. Autistic Spectrum Disorders), and the study of functional connectivity.
Finally, key challenges are addressed, including signal quality, data variability, model interpretability, scalability, and ethical concerns. The design, validation, and clinical deployment of AI models are discussed with a focus on how these methods can provide researchers and clinicians with clear, actionable information about brain performance, disease progression, and personalized interventions, supporting the development of precision neuroscience and improved brain health.
Marçal Mora-Cantallops
Marçal Mora-Cantallops is an Associate Professor in the Computer Science Department atthe University of Alcalá, in Alcalá de Henares, Spain. He is part of the Biomedical Data
Science and Engineering group within the IRyCIS research institute (part of the Ramón y
Cajal Hospital in Madrid, Spain), which aims to research and apply computational methodsto all kinds of problems in biomedical contexts. As a researcher, he focuses on machine learning, social network analysis, data science, ethical AI, and game studies, havingauthored and published multiple related articles in the past few years. He also coordinatedthe Trustworthy AI Erasmus+ project and is currently coordinating the AIMS (AI for Medical Students) Erasmus+ project.
The title of the presentation: Preparing Future Clinicians for AI-Driven Biomedicine: the Educational Approach from the AIMS Erasmus+ Project
Abstract
Advances in artificial intelligence (AI), data-driven modelling, and digital technologies are increasingly influencing biomedical research and clinical practice. While these developments are reshaping diagnostic, monitoring, and therapeutic approaches, medical education is still adapting to prepare future clinicians to critically engage with such technologies.
This talk presents the AIMS (Artificial Intelligence for Medical Students) Erasmus+ cooperation project, which is currently in progress and focuses on supporting the responsible integration of AI into undergraduate medical education. Rather than addressing technical implementation alone, AIMS responds to a growing educational need: enabling medical students, educators, and academic leaders to understand, evaluate, and ethically use AI-supported tools in complex biomedical contexts.
At its current stage, AIMS is developing the Synergy Matrix, a structured mapping tool that links AI applications and data-driven technologies to existing medical curricula. This work is informed by consultations with medical educators, AI specialists, and academic leaders, and is designed to reflect real curricular constraints and clinical relevance. The project is also laying the foundations for an AI Competence Framework and related teaching resources that will support educators in embedding AI concepts meaningfully into medical training.
The presentation will focus on the AIMS approach, outlining the project’s methodology, early design choices, and underlying educational assumptions. Rather than presenting final results, the session aims to share work in progress and actively invite feedback from the academic and research community. Input from experts working at the intersection of biomedicine, physics, and technology will be used to refine the Synergy Matrix and inform subsequent project stages.
Science and Engineering group within the IRyCIS research institute (part of the Ramón y
Cajal Hospital in Madrid, Spain), which aims to research and apply computational methodsto all kinds of problems in biomedical contexts. As a researcher, he focuses on machine learning, social network analysis, data science, ethical AI, and game studies, havingauthored and published multiple related articles in the past few years. He also coordinatedthe Trustworthy AI Erasmus+ project and is currently coordinating the AIMS (AI for Medical Students) Erasmus+ project.
The title of the presentation: Preparing Future Clinicians for AI-Driven Biomedicine: the Educational Approach from the AIMS Erasmus+ Project
Abstract
Advances in artificial intelligence (AI), data-driven modelling, and digital technologies are increasingly influencing biomedical research and clinical practice. While these developments are reshaping diagnostic, monitoring, and therapeutic approaches, medical education is still adapting to prepare future clinicians to critically engage with such technologies.
This talk presents the AIMS (Artificial Intelligence for Medical Students) Erasmus+ cooperation project, which is currently in progress and focuses on supporting the responsible integration of AI into undergraduate medical education. Rather than addressing technical implementation alone, AIMS responds to a growing educational need: enabling medical students, educators, and academic leaders to understand, evaluate, and ethically use AI-supported tools in complex biomedical contexts.
At its current stage, AIMS is developing the Synergy Matrix, a structured mapping tool that links AI applications and data-driven technologies to existing medical curricula. This work is informed by consultations with medical educators, AI specialists, and academic leaders, and is designed to reflect real curricular constraints and clinical relevance. The project is also laying the foundations for an AI Competence Framework and related teaching resources that will support educators in embedding AI concepts meaningfully into medical training.
The presentation will focus on the AIMS approach, outlining the project’s methodology, early design choices, and underlying educational assumptions. Rather than presenting final results, the session aims to share work in progress and actively invite feedback from the academic and research community. Input from experts working at the intersection of biomedicine, physics, and technology will be used to refine the Synergy Matrix and inform subsequent project stages.