Research Group for Applied Software Engineering
Forschungsgruppe für Angewandte Softwaretechnik

Maximilian Kapsecker

Office: 01.07.037
E-Mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
Phone: +49 (89) 289 - 18239

Technische Universität München
Institut für Informatik I1
Boltzmannstraße 3
85748 Garching b. München, Germany
Office hours by appointment—please contact me via mail.

 

Hello, it's Max! I am affiliated with the Digital Health Group (TUM) and the Institute for Digital Medicine (UKB).

Research Fields

My research area is at the intersection of mobile software engineering, machine learning, and medicine. Specifically, I am interested in using mobile sensor data to detect anomalies in an unsupervised manner to infer medical conditions. The nature of this sensitive data requires consideration of security aspects, such as secure aggregation, differential privacy, and federated learning.

#federated-representational-learning #mobile-sensor-data #medical-applications

Projects

  • Federated representational learning for the detection of anomalies in electrocardiogram signals
  • Deep embedding of biosignals for the detection of anomalies
  • Software Campus and Zeiss: Detection of vision impairment with mobile devices
  • MRI - Telemonitoring of athletes and patients (iPraktikum SS21 - WS21/22 - SS22)
  • Solgenium - Verifying workflows in a clinical environment (iPraktikum SS21)
  • footpower - Mobile measurement of the human leg (iPraktikum WS 20/21)

Publications

2023

  • Starnecker, F., Reimer, L. M., Nissen, L., Jovanović, M., Kapsecker, M., Rospleszcz, S., ... & Schunkert, H. (2023). Guideline-Based Cardiovascular Risk Assessment Delivered by an mHealth App: Development Study. JMIR cardio, 7(1), e50813.
  • McRae, H. L., Kahl, F., Kapsecker, M., Rühl, H., Jonas, S. M., & Pötzsch, B. (2023). Evaluation of an Explainable Tree-Based AI Model for Thrombophilia Diagnosis and Thrombosis Risk Stratification. Blood142, 2300.
  • Kristof, F., Kapsecker, M., Nissen, L., Brimicombe, J., Cowie, M., Ding, Z., ... & Charlton, P. H. (2023). QRS detection in single-lead, telehealth electrocardiogram signals: benchmarking open-source algorithms. medRxiv, 2023-11.
  • Kapsecker, M., Nugraha, D. N., Weinhuber, C., Lane, N., & Jonas, S. M. (2023). Federated Learning with Swift: An Extension of Flower and Performance Evaluation. SoftwareX24, 101533.
  • Nissen, L., Hübner, J., Klinker, J., Kapsecker, M., Leube, A., Schneckenburger, M., & Jonas, S. M. (2023). Towards Preventing Gaps in Health Care Systems through Smartphone Use: Analysis of ARKit for Accurate Measurement of Facial Distances in Different Angles. Sensors, 23(9), 4486.

2022

  • Kapsecker, M., Nissen, L., Weinhuber, C., & Jonas, S. (2022, September). Künstliche Intelligenz zur Erkennung von Zustandsänderungen in hochdimensionalen Patientendaten. Künstliche Intelligenz in der Medizin, Heft 3, 106.
  • Kapsecker, M., Osterlehner, S., & Jonas, S. M. (2022, July). Analysis of Mobile Typing Characteristics in the Light of Cognition. In 2022 IEEE International Conference on Digital Health (ICDH) (pp. 87-95). IEEE.
  • Kapsecker, M., Strobel, B., & Jonas, S. (2022). SAFMA: Secure Aggregation Framework for mHealth Applications. 67. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V. (GMDS)
  • Reimer, L. M., Kapsecker, M., Fukushima, T., & Jonas, S. M. (2022). Evaluating 3D Human Motion Capture on Mobile Devices. Applied Sciences12(10), 4806.

Theses

Open - all titles are preliminary and require further discussion

  • Posterior collapse in variational autoencoders for time series data
  • Real-time detection of cardiac anomalies in mobile environments
  • Integration of federated learning in a running healthcare system

 

2023

  • Ismail Ulas Bayram - Architecture of a Federated Learning System to Detect Anomalies in Mobile Health Records
  • Devansh Sharma - Detecting Facial Micro-Gestures as a Predictor of Visual Impairment
  • Sonja Krafft - Investigating Disentanglement in Electrocardiogram Representations: Assessing Latent Factors and Similarity Measures
  • Léon Friedmann - A Mobile Framework for Sleep Record Analysis and Outlier Detection Providing Indicators for Mental Health
  • Richard Pfannenstiel - Automatic Camera-based Assessment of the Short Physical Performance Battery on Mobile Devices
  • Atakan Özcan - Location-Based Mental Health Analysis on Mobile Devices
  • Junpeng Chen - A Privacy-Preserving Machine Learning Approach for Mental Health Assessment using Communication Patterns
  • Nikita Charushnikov - PeakSwift - A Swift Package for the Detection of QRS Complexes in Single-Lead Electrocardiogram Signals
  • Fedor Shvetsov - Implementation of a System to Analyse Factors Affecting Vision while using Digital Screens
     

2022

  • Alina Usova - Semi-supervised Labelling System of Electrocardiogram Signals for the Detection of Cardiac Anomalies
  • Carmen Berndt - Personalized Mobile Device Disglycemia Prediction based on Electrocardiogram Signals
  • Linus Salzmann - Efficient Compression of Electrocardiogram Signals on Mobile Devices using Autoencoders
  • Julian Kretzschmar - Autoencoders for the Personalized Detection of Electrocardiogram Baseline Drifts on Mobile Devices
  • Florian Kristof - Assessment of Electrocardiogram Beat Detectors Using Synthetic and Real-World Data
  • Daniel Nugraha - A Prototype Implementation of a Mobile Federated Learning Framework
  • Ramona Eckert - Development of a Mobile Application for Automated Functional Movement Analysis
  • Cornelius Born - Predicting Dysglycemia for Individuals with Diabetes Mellitus from Wearable Data
  • Simon Osterlehner - Development of a privacy-preserving keyboard application for the detection of dementia
  • Benedikt Strobel - Implementation of a Scalable Secure Aggregation Framework for Mobile Applications
  • Sebastien Letzelter - Implementing a Mobile Application for Detecting Vision Impairments

 

2021

  • Simone Schmidpeter - Federated Learning for Anomaly Detection in Electrocardiograms using Clustering
  • Stephan Schmiedmayer - Implementation of a Secure Aggregation Protocol for Mobile Devices to Model Health Related Parameters
  • Victor Dzhagatspanyan - Comparison of generative models for the synthetic generation of electrocardiogram data
  • Ernst Pappenheim - Conception and Implementation of a Software Tool to Optimize and Standardize Data Management in Personnel Planning in the Clinical Environment