The accepted contributions highlight our group’s interdisciplinary research at the intersection of artificial intelligence, digital health, computer vision, wearable sensing, and clinical epileptology. Our work focuses on the development of digital biomarkers and AI-based analysis methods to improve the monitoring, diagnosis, and understanding of epilepsy.

Research Topics

The presented studies cover several emerging directions in AI-driven epilepsy research, including:

• Contactless measurement of vital parameters using camera-based sensing technologies
• 3D human pose estimation for video-EEG monitoring to analyze seizure-related body movements
• Eye-tracking metrics for cognitive assessment in epilepsy, including contextualized analysis of visual search strategies
• Effects of anti-seizure medication on eye-tracking parameters in patients with epilepsy
• Eye tracking during the Trail Making Test and its association with depressive symptoms
• Secure and trustworthy AI models for seizure detection using wearable devices
• Electrocardiographic changes under cenobamate therapy based on extended ECG recordings

These contributions reflect ongoing collaborations between computer scientists, physicists, neurologists, and clinical researchers at the University Hospital Bonn and partner institutions.

A large part of the work has been carried out by PhD students and early-career researchers, demonstrating the strong integration of young scientists in our research activities.

We look forward to presenting our work in Würzburg and discussing these results with colleagues from the international epilepsy research and digital medicine community.

@conference{simsek2026a,

title = {Kontaktloses kamerabasiertes Messen von Vitalparametern},

author = {Koray Simsek and Johannes Müllers and Rainer Surges and Björn Krüger},

year  = {2026},

date = {2026-06-13},

urldate = {2026-06-13},

booktitle = {64. Jahrestagung der Deutschen Gesellschaft für Epileptologie},

keywords = {},

pubstate = {forthcoming},

tppubtype = {conference}

}

@conference{nokey,

title = {Kontextualisierte Eye-Tracking-Metriken zur Charakterisierung von Suchstrategien bei Personen mit Epilepsie und Kontrollen},

author = {Anna Jansen and Melissa Steininger and Sarah Al-Haj Mustafa and Nataly Bouzan and Rainer Surges and Christoph Helmstaedter and Randi von Wrede and Björn Krüger},

year  = {2026},

date = {2026-06-13},

urldate = {2026-06-13},

booktitle = {64. Jahrestagung der Deutschen Gesellschaft für Epileptologie},

keywords = {},

pubstate = {forthcoming},

tppubtype = {conference}

}

@conference{gress2026a,

title = {Anforderungen an sichere KI-Modelle zur Anfallsdetektion mit Wearables in der Epileptologie},

author = {Hannah Greß and Nazila Ahmadi Daryakenari and Christian Bungartz and Felix Viola and Marco Markwald and Gabriela Brüll and Uttam Kumar and Marc Ohm and Rainer Surges and Michael Meier and Elena Demidova and Björn Krüger},

year  = {2026},

date = {2026-06-13},

urldate = {2026-06-13},

booktitle = {64. Jahrestagung der Deutschen Gesellschaft für Epileptologie},

keywords = {},

pubstate = {forthcoming},

tppubtype = {conference}

}

@conference{vetter2026,

title = {Kontaktlose 3D-Human-Pose-Estimation im Video-EEG-Monitoring von Epilepsiepatienten},

author = {Jonas Vetter and Johannes Müllers and Federico Spurio and Rainer Surges and Juergen Gall and Björn Krüger},

year  = {2026},

date = {2026-06-13},

urldate = {2026-06-13},

booktitle = {64. Jahrestagung der Deutschen Gesellschaft für Epileptologie},

keywords = {},

pubstate = {forthcoming},

tppubtype = {conference}

}

@conference{Pukropski2026a,

title = {Elektrokardiographische Veränderungen unter Cenobamat – eine retrospektive Prä-Post-Analyse aus verlängerten EKG-Ableitungen bei Patient*innen mit Epilepsie},

author = {Jan Pukropski and Lisa Keßler and Randi von Wrede and Rainer Surges and Björn Krüger},

year  = {2026},

date = {2026-06-13},

urldate = {2026-06-13},

booktitle = {64. Jahrestagung der Deutschen Gesellschaft für Epileptologie},

keywords = {},

pubstate = {forthcoming},

tppubtype = {conference}

}

@conference{steininger2026c,

title = {Zusammenhänge zwischen Anfallssuppressiva und kontextualisierten Eye-Tracking-Metriken bei Menschen mit Epilepsie},

author = {Melissa Steininger and Anna Jansen and Sarah Al-Haj Mustafa and Nataly Bouzan and Rainer Surges and Christoph Helmstaedter and Randi von Wrede and Björn Krüger},

year  = {2026},

date = {2026-06-13},

urldate = {2026-06-13},

booktitle = {64. Jahrestagung der Deutschen Gesellschaft für Epileptologie},

keywords = {},

pubstate = {forthcoming},

tppubtype = {conference}

}

@conference{mustafa2026a,

title = {Wer suchet, der findet: Eye Tracking beim Trail Making Test bei Epilepsie – Zusammenhänge mit depressiver Symptomatik},

author = {Sarah Al-Haj Mustafa and Anna Jansen and Melissa Steininger and Johannes Müllers and Rainer Surges and Christoph Helmstaedter and Björn Krüger and Randi von Wrede},

year  = {2026},

date = {2026-06-13},

urldate = {2026-06-13},

booktitle = {64. Jahrestagung der Deutschen Gesellschaft für Epileptologie},

keywords = {},

pubstate = {forthcoming},

tppubtype = {conference}

}

The paper “From Steps to Sentiments: Cross-Domain Transfer Learning for Activity-Based Emotion Detection in Wearable IoT Systems” explores how machine learning models trained for activity recognition can be adapted to detect emotional states. Using a cross-domain transfer learning approach, the study demonstrates how knowledge learned from physical activity data can be transferred to emotion recognition tasks.

Wearable devices continuously capture movement patterns, step dynamics, and physical activity signals, which provide valuable information about human behavior in everyday environments. By leveraging these signals, the proposed approach opens new opportunities for non-invasive emotion monitoring using wearable IoT systems.

Such technologies have potential applications in mental health monitoring, digital phenotyping, and personalized healthcare, where unobtrusive sensing and intelligent data analysis can support early detection of behavioral and emotional changes.

This publication also continues a long-standing research collaboration between Qaiser Riaz and Björn Krüger, focusing on machine learning methods for human movement analysis and wearable sensor systems.

The full publication can be accessed via IEEE Xplore:
https://ieeexplore.ieee.org/document/11404152

@article{Imran2026a,

title = {From Steps to Sentiments: Cross-Domain Transfer Learning for Activity-Based Emotion Detection in Wearable IoT Systems},

author = {Hamza Ali Imran and Qaiser Riaz and Kiran Hamza and Shaida Muhammad and Björn Krüger},

doi = {10.1109/JIOT.2026.3666469},

year  = {2026},

date = {2026-02-20},

urldate = {2026-02-20},

journal = {IEEE Internet of Things Journal},

pages = {1-1},

abstract = {Context-aware, gait-based sentiment analysis and emotion perception is an emerging research area within Internet of Things (IoT), aiming to make smart systems more intuitive and responsive. Recognizing emotions from wearable inertial sensor data is challenging due to subtle and compound emotional cues, variability across individuals and contexts, and limited, imbalanced datasets. To address these challenges, we propose Jazbat-Net, a lightweight neural network that leverages Transfer Learning (TL). The model is first trained on a large-scale, publicly available multi-activity dataset collected using wearable inertial sensors, and then retrained on a multi-class emotion dataset, effectively transferring knowledge from the pretraining phase. We evaluate Jazbat-Net with and without TL, across both smartwatch and smartphone based data, and for input dimensions ranging from 1D to 6D. The best results are achieved when pretrained on smartphone-based activity data and retrained on smartphone-based emotion data using a 1D input size. The proposed model attains an average classification accuracy of 95%, with a precision score of 95%, a recall score of 97%, and an F1-score of 96%. Moreover, Jazbat-Net achieves a low theoretical time complexity and requires only ≈ 6.96 M Multiply–Accumulate Operations (MACs), which is about 95% fewer computations than the previous State-of-the-Art (SOTA) model. Its space complexity is also low, with a model size of only ≈ 110 KB and peak activation memory of ≈ 0.35 MB. On-device evaluation on a Xiaomi 13T smartphone demonstrates that Jazbat-Net achieves a median inference latency of only ≈ 90.96 ms with a TFLite 32-bit floating point precision (FP32) model size of just ≈ 0.158 MB, making it ≈ 20× smaller and ≈ 20% faster than the previous SOTA model while maintaining comparable accuracy.

},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

The accepted contributions are:

• Linking Higher-Level Eye Tracking Metrics to High-Impact Antiseizure Medication in Epilepsy Patients
• Higher-Level Eye Tracking Metrics Reveal Search Behaviour Differences in Persons with Epilepsy vs. Healthy Controls

Anna Jansen and Melissa Steininger will present this work at the conference.

We look forward to continuing our research on digital health and AI in epilepsy in the coming year and wish everyone a restful holiday season.

The symposium brought together an international community of researchers, offering a rich setting for scientific exchange. Prof. Krüger particularly valued the discussions that followed his keynote, which touched on both technical developments and broader questions surrounding trustworthy AI in medicine.

The event was organized by Shadi Albarqouni, whose efforts created an inspiring environment for interdisciplinary dialogue.

In a short talk, Kristian outlined the project’s goals and technical approach. A particular highlight was the drilling simulation demo with a haptic arm, which provided highly realistic tactile feedback and sparked strong interest among visitors.

VIRTOSHA fits seamlessly into the HealthTech.NRW focus area of high-tech surgery – an environment where robots, lasers, and VR are already part of everyday clinical practice in NRW, significantly enhancing patient safety and treatment quality.

Many thanks to all visitors for the inspiring conversations and positive feedback!

His contribution highlighted ongoing projects ranging from wearable-based motion analysis—using lightweight models for activity and emotion recognition—to eye-tracking studies in epilepsy that investigate how cognitive changes and medication side effects manifest in eye movements.

The event offered a valuable opportunity for exchange with fellow panelists Stephen Lee (College of Medicine, University of Saskatchewan) and Raymond Ng (Data Science Institute / AI & Health Network, University of British Columbia) and Daniel Fuller (University of Saskatchewan) who did an excellent job in moderation and organization.

The discussion underscored a central theme of his work: AI in healthcare is not only about accuracy—its true impact lies in improving quality of life, supporting clinicians, and empowering patients.

In his talk, he emphasized three key insights drawn from these experiences:

1. Peer review is unpredictable. Individual reviews often reflect randomness; they should not define one’s self-worth.
2. Failures can spark better ideas. Rejections frequently lead to rethinking, refining, and ultimately creating stronger work.
3. Career detours are valuable. Unexpected paths can lead to the places where one truly belongs.

The event featured a series of inspiring contributions from speakers who shared their own stories of setback, learning, and resilience.

Accepted Contributions – Dreiländertagung Epilepsie 2025 in Salzburg:
Digital Transformation of Blood Sample Management
EpiEye – Effects of Seizure-Suppressive Medications on Eye Movements and Autonomic Changes in Epilepsy
Establishing an Epileptological Teleconsultation Between Siegen Hospital and University Hospital Bonn
ANNE – Eye-Tracking for Detecting Side Effects in Epilepsy
Semiological Characteristics of Temporal Lobe Epilepsy with GAD65 Antibodies

Accepted Contributions – International Conference on Artificial Intelligence in Epilepsy and Other Neurological Disorders 2025:
Eye-Tracking Reveals Search Behaviour in Epilepsy Patients
Prediction Models on Eye-Tracking Data in Epilepsy

Accepted Contribution – Wissenschaftlichen Tagung Autismus-Spektrum (WTAS)
Multi-Stream Analysis for Robust Head Gesture Classification in Natural Social Interaction: Leveraging High-Resolution Sensor Data for Optimized Visual Classification

We sincerely thank all contributors for their dedication and hard work. We look forward to engaging discussions, inspiring exchanges, and exciting collaborations at these upcoming conferences!

Wishing everyone a successful and inspiring year ahead!

The program of the workshop can be seen here.