Arlene John is an Assistant Professor at the Biomedical Signals and Systems (BSS) group. Arlene completed her Bachelor’s degree in Electrical and Electronics Engineering at the National Institute of Technology, Calicut, India, in 2017. In the summer of 2016, she worked as a Research Intern at the Indian Institute of Science, Bangalore, India. After graduation, she joined Bosch India Ltd. as a Project Manager in Technical Sales and later worked in engineering and strategy development for hybrid electric vehicles.
She obtained her Ph.D. from the School of Electrical and Electronic Engineering, University College Dublin, Ireland, in 2022, where her research focused on developing data fusion frameworks for wearable health monitoring devices. From March to June 2019, she was a Senior Visiting Researcher at Beijing University of Technology, China, and in 2021, she worked as a Machine Learning Intern at Qualcomm, Cork, Ireland. After her Ph.D., she joined ASML Netherlands B.V. as a Machine Learning Mathematics Engineer, where she worked until April 2023.
Her research interests include explainable AI, multisensor data fusion, and biomedical signal processing and modeling.
Expertise
Computer Science
- Internet of Things
- Convolutional Neural Network
- Models
- Detection
- Noise-to-Signal Ratio
Medicine and Dentistry
- Postoperative Complication
- Breathing Rate
Engineering
- Atrial Fibrillation
Organisations
My research is at the intersection of explainable artificial intelligence and life sciences, through which I aim to shape the future of sustainable and equitable digital healthcare.
Most of the code and models I develop are published on my github: https://github.com/arlenejohn
I also curate a podcast on Spotify to better communicate my research, where every episode is narrated by an advanced language model: https://open.spotify.com/show/1JWZF0jL5KCnQeIoLWcSO9
The Springer Nature book based on my research on multisensor data fusion: Deep Learning and Signal-Processing Methods for Multisensor Data Fusion: Applications to Ambulatory Health Monitoring | Springer Nature Link
Publications
2026
2025
Research profiles
I coordinate the Medical Sensors and Measurements course in the Biomedical Engineering Bachelor’s program (Quarter 3).
I lecture in the Multimodal Machine Learning course in the Computer Science Master's program (Quarter 3).
Affiliated study programs
Courses academic year 2025/2026
Courses in the current academic year are added at the moment they are finalised in the Osiris system. Therefore it is possible that the list is not yet complete for the whole academic year.
Courses academic year 2024/2025
Granted Projects:
There is a great need for a more personalized approach in healthcare and smart monitoring solutions to revolutionize healthcare and relocate care from the clinical to the home setting. In RECENTRE, WP2, we develop dynamic risk profiles based on multivariate data for the added risk of lifestyle on health outcomes over time, particularly within late effects after cancer and obesity treatment.
Neurological conditions in newborns can cause lifelong disabilities if not diagnosed early. Around 15 million babies are born prematurely each year, many at risk of seizures and brain injuries. The INSIGHT project will develop explainable AI algorithms to detect critical neurological events using graph neural networks (GNN). A GNN-based model is proposed to incorporate interdependencies between EEG leads. Explainable AI development will aid in decision interpretation for diagnoses and to generate automated summary reports from EEG recordings. The AI model will be integrated into NeuroBell’s portable EEG device, with an intuitive dashboard to assist clinicians in decision-making. The system will be tested in real-world clinical settings with 10 neonates, ensuring usability and effectiveness.
Heart failure is an escalating global health challenge, affecting over 64 million people worldwide. Despite advancements in guideline-directed medical therapy (GDMT) that significantly reduce mortality and hospitalizations, many patients still do not receive optimal medication regimens or dosages. The project is a collaborative effort uniting multiple stakeholders, including healthcare providers, researchers, and (biomedical) engineers. This consortium aims to evaluate the impact of a cutting-edge digital care intervention designed to streamline medication optimization. By leveraging a robust remote monitoring infrastructure, this approach seeks to make the process more efficient, scalable, and accessible, while focusing on improving critical clinical outcomes.
CARE
The CARE project aims to develop a proof-of-concept for the early detection of complications in neonates/children (0-2 years) who have undergone colorectal surgery through unobtrusive continuous monitoring of vital signs. In the broader picture, this can lead to the development of enhanced recovery after surgery (ERAS) programs in pediatric patients, which is lagging behind compared to the adult population. In adults, ERAS has proven to improve patient outcomes through early mobilization and early discharge, but it relies on obtrusive tests (blood tests, imaging, etc. not suitable for children) to ensure patient safety. Continuous monitoring will help narrow this gap in pediatric care by using wireless sensors to measure vitals. Our goal is to correlate trends in patient vitals, measured through wireless sensors, to complications or normal recovery.
ExplAIn-HF
Heart failure (HF) is an urgent problem in the Netherlands, where with current treatment 33,000 patients are admitted to hospital annually.In the current care pathway, healthcare providers often make decisions based on limited data and subjective interpretation. There is a lack of uniform decision-making and a lack of transparent predictive models that help healthcare providers to properly substantiate decisions. We aim to develop explainable (eXplainable) AI models (XAI) for forecasting HF events using a wearable. This approach allows for the collection of more detailed data and the identification of complex patterns relevant for predicting deterioration in chronic and ambulatory HF patients. The data is collected non-invasively, improving the current care pathway, which is primarily accessible to digitally literate patients.
EMERALD
While lifestyle is recognized as an important topic in chronic disease care, lack of time and knowledge are important barriers for implementation. Traditional face-to-face interventions report moderate effects and are time- and resource-intensive. As the healthcare systems face mounting pressure from workforce shortages, and rising costs, highlighting the need for scalable, preventive, and patient-centered approaches that can be accomplished with eHealth technologies. Large language models (LLMs) offer an innovative and effective solution towards highly personalized support by generating human-like, context-aware dialogue that translates complex medical information into clear, supportive communication for daily self-management and lifestyle change. However, concerns regarding data privacy, ethical considerations, factual accuracy and potential bias are recognized for applying LLMs in healthcare and hinder its implementation in healthcare pathways. To ensure that patients will receive guidance that is in line with clinical guidelines and thus safe, we aim to build our own behavior-change LLM fine-tuned on chronic-disease populations, i.e., to diabetes type 2 (DM2), cardiovascular disease (CVD) and knee osteoarthritis (KOA) that can be implanted into healthcare pathways. This will allow us to optimize safety, domain specificity, thereby increasing adherence, improving clinical outcomes, and reducing long-term healthcare burden.
Unsuccessful project applications:
If you are interested in knowing more about any of the unsuccessful grant applications listed below, and would like to collaborate on the same topics, write to me.
1. Continuous monitoring of patients after colorectal surgery at hospitals and homes.
2. Quantifying cardiac remodelling during pregnancy for early detection of pre-eclampsia and hypertension.
3. Boosting Long-term health Outcomes and Optimising self Management forprevention through digital health and holistic support for children.
Address
University of Twente
Horst Complex (building no. 20), room ZH210
De Horst 2
7522 LW Enschede
Netherlands
University of Twente
Horst Complex ZH210
P.O. Box 217
7500 AE Enschede
Netherlands
Scan the QR code or