What we do
Understanding Cancer Pain
Pain is one of the most common and burdensome symptoms experienced by people with cancer. It can affect movement, sleep, appetite, mood, independence, social participation and overall quality of life. Despite major advances in cancer care, pain is still not always recognised or treated adequately. A substantial proportion of patients continue to experience undertreated pain.
One reason for this is that cancer-related pain is difficult to fully understand from the outside. Pain is not only a symptom that can be scored or observed; it is also a personal and lived experience. For each patient, pain has its own pattern, meaning and impact. It may be constant or unpredictable, sharp or dull, manageable at one moment and overwhelming at another. It can also be closely connected to fatigue, emotional distress, uncertainty, treatment side effects and changes in daily life.
In clinical practice, pain is usually assessed during consultations or captured with a numerical score. These methods are important, but they can only reflect part of the patient’s experience. A single score may indicate how intense pain feels at a specific moment, but it does not necessarily show how pain affects daily activities, emotional well-being, comfort, independence or the need for support. Furthermore, current practices depend on what a patient is able to express during that particular conversation.
This challenge is especially relevant in outpatient cancer care. Patients often spend most of their time at home, between scheduled appointments. During these periods, pain may change, increase or interfere with daily life in ways that remain largely invisible to the care team. As a result, opportunities to recognise concerns, start conversations or adapt support may be missed.
Our vision
The SENSAI Pain project aims to improve how cancer-related pain is understood, communicated and managed in oncology care. We are developing a mobile application that helps patients share their pain experience between hospital visits. The application combines patients’ own pain reports with short questions about mood and daily functioning. It also uses AI to analyse facial expressions and voice when patients describe their pain. Together, this can provide a fuller picture of pain in daily life. With SENSAI Pain, we aim to support more meaningful conversations between patients and healthcare professionals and contribute to pain care that better reflects what patients experience.
A collaborative and ethical approach
SENSAI is built through close collaboration between researchers, healthcare specialists, engineers, and patients. We follow a human-centred design process in which feedback from end-users guides every step of development. Their perspectives help shape both the technology and the way it may be used in care. This approach ensures the tool fits naturally into daily care, respects privacy, and remains understandable and trustworthy for both patients and clinicians.
Expected impact
By combining state-of-the-art technology with clinical expertise, SENSAI aims to empower patients to reflect on their pain, give doctors clearer insight into its severity and pattern, and support more personalised and timely pain management. Ultimately, the project strives to improve comfort and communication throughout the cancer journey.Start of the project
Unlike many technical projects, SENSAI begins with people rather than algorithms. Early interviews with oncologists have shown how complex it can be to interpret pain from patients’ descriptions and how easily important cues are missed. These insights guide the design of the tool, ensuring that technology supports – rather than replaces – the personal interaction between patients and clinicians. Interviews with patients are currently underway to explore their experiences with cancer pain and their needs for communication and self-management support.The SENSAI application
The SENSAI app enables patients to record short, guided sessions in which they describe their pain and answer a few brief questions about pain intensity and mood. The app captures both facial expressions and voice characteristics while the patient speaks. These recordings are processed by the AI model to identify patterns associated with different pain levels. After analysis, the app will offer patients a simple visual overview of their pain experiences, helping them recognise changes and communicate more effectively with their healthcare team.The SENSAI cancer pain database
To train and validate the AI model, SENSAI is establishing a secure multimodal cancer pain database at Erasmus MC. This data collection study is registered at ClinicalTrials.gov(NCT07262632). This database contains synchronised recordings of facial video, voice audio, and self-reported pain and mood data collected in a controlled clinical environment. All data are handled according to strict ethical and privacy standards. In the future, anonymised data will be made available to collaborating researchers, supporting transparency and further innovation in automatic pain assessment.The SENSAI AI model
The SENSAI AI model integrates information from facial and vocal features with patients’ self-reported pain scores to estimate pain intensity levels. Its design focuses on transparency and explainability, ensuring that clinicians can understand and trust how the model reaches its conclusions. Future developments will include analysis of spoken content and emotional tone to better capture the complex and multidimensional nature of cancer pain.Internal collaborations
The SENSAI project is carried out in the Department of Medical Oncology at Erasmus MC.
External collaborations
The project is carried out in close collaboration with the Multi-Actor Systems group at Delft University of Technology (TU Delft), where several members of the research team are affiliated. This partnership contributes expertise in human-centred design and state-of-the-art artificial intelligence.
Furthermore, the mobile application is developed together with Innovattic, a software company based in Delft, the Netherlands. See Innovattic.com for more information.
  |
Kamsteeg, M. J. (2025). Towards an AI-Empowered Multimodal pain Assessment Tool for Cancer-Related pain [Master Thesis Technical Medicine, University of Technology Delft]. https://repository.tudelft.nl/record/uuid:efaadcb3-03fa-4f8e-ac87-33f5098f06af
Kamsteeg M. J. (2025, October 9). Perspectieven van oncologen op pijnbeoordeling en AI-ondersteunde automatische pijnbeoordeling bij kanker: een explorerende interviewstudie [Conference presentation] Nederlands Vlaamse Wetenschapsdagen Palliatieve Zorg
Kamsteeg M.J. (2025, November 12-14). Towards an Artificial Intelligence Based Multimodal Pain Assessment Tool for Cancer-Related Pain: Foundations from the SENSAI Project [Conference poster] ESMO Artificial Intelligence & Digital Oncology
Kamsteeg M.J. (2026, April 9-10). Oncologists’ Perspectives on Pain Assessment and AI-Supported Automatic Pain Assessment in Cancer: An Exploratory Interview Study [Conference presentation] Cancer Research Retreat
Kamsteeg M.J., Algera F.L., Van der Rijt C.D.D., Mulder M., Torkamaan H. (2026, June) From Prototype to Data Quality: Usability Findings for a Patient-Facing AI-Empowered Mobile Application for Automatic Cancer Pain Assessment [Conference proceeding] IEEE Computer-Based Medical Systems
Coordinating researcher
m.kamsteeg@erasmusmc.nl LinkedIn
Prof. Karin van der Rijt – Medical Oncologist & Professor Palliative Care – Erasmus MC Cancer Institute
Dr. Wendy Oldenmenger – Assistant Professor Palliative Care – Erasmus MC Cancer Institute
Dr. ir. Helma Torkamaan – Assistant Professor AI for Health Systems – TU Delft (see torkamaan.eu)
Dr. ir. Mark Mulder – Medical Oncologist – Erasmus MC Cancer Institute
