Bio

I am a postdoctoral researcher at the Mignot Lab in Stanford University. My background is in biomedical engineering, signal processing and machine learning. I obtained a BSc and MSc degree from KU Leuven in 2017 and 2019, respectively. After this, I performed a research internship at École Polytechnique Fédérale De Lausanne in the Neuroengineering Lab. I did my PhD (2020-2024) under the supervision of Prof. Maarten De Vos, focusing on automated sleep staging using electroencephalography and polysomnography signals. During my PhD, I also performed a research stay at the University of Cambridge (van der Schaar lab) and an internship at Microsoft Research (in the Brain-Computer Interfaces project).

During my postdoc at the Mignot Lab, I aim to use large sleep datasets to find early markers of depression or other disorders related to sleep. My main research interest lies in the intersection between AI and neuroscience, and using AI for neuroscientific applications.

Professional Education

  • Doctor of Philosophy, Katholieke Universiteit Leuven (2024)
  • Master of Science, Katholieke Universiteit Leuven (2019)
  • Bachelor of Science, Katholieke Universiteit Leuven (2017)
  • PhD, KU Leuven, Doctor of Engineering Science (Electrical Engineering) (2024)
  • MSc, KU Leuven, Engineering Science (Biomedical Engineering) (2019)
  • BSc, KU Leuven, Engineering Science (Computer Science) (2017)

Stanford Advisors

Contact

  • Academic
    heremans@stanford.edu
    University - Scholar Department: Psychiatry and Behavioral Sciences Position: Postdoctoral Scholar

Additional Info

All Publications

  • Personalization of Automatic Sleep Scoring: How Best to Adapt Models to Personal Domains in Wearable EEG IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS Lorenzen, K. P., Heremans, E. M., de Vos, M., Mikkelsen, K. B. 2024; 28 (10): 5804-5815

    Abstract

    Wearable EEG enables us to capture large amounts of high-quality sleep data for diagnostic purposes. To make full use of this capacity we need high-performance automatic sleep scoring models. To this end, it has been noted that domain mismatch between recording equipment can be considerable, e.g. PSG to wearable EEG, but a previously observed benefit from personalizing models to individual subjects further indicates a personal domain in sleep EEG. In this work, we have investigated the extent of such a personal domain in wearable EEG, and review supervised and unsupervised approaches to personalization as found in the literature. We investigated the personalization effect of the unsupervised Adversarial Domain Adaptation and implemented an unsupervised method based on statistics alignment. No beneficial personalization effect was observed using these unsupervised methods. We find that supervised personalization leads to a substantial performance improvement on the target subject ranging from 15% Cohen's Kappa for subjects with poor performance ( ) to roughly 2% on subjects with high performance ( ). This improvement was present for models trained on both small and large data sets, indicating that even high-performance models benefit from supervised personalization. We found that this personalization can be beneficially regularized using Kullback-Leibler regularization, leading to lower variance with negligible cost to improvement. Based on the experiments, we recommend model personalization using Kullback-Leibler regularization.

    View details for DOI 10.1109/JBHI.2024.3409165

    View details for Web of Science ID 001329782300020

    View details for PubMedID 38833404

https://orcid.org/0000-0001-6860-2356