Info Support is to support the Netherlands Organisation for Scientific Research (NWO) in conducting research into and developing advanced artificial intelligence (AI). The AI models should make care in oncology pathways more personalised, as well as relieve the burden of care.
The collaboration forms part of the NWO’s PersOn programme. Within this programme, the NWO is using artificial intelligence (AI) to analyse all available data on a cancer patient and the type of cancer they have. The aim is to predict the expected outcome of various treatment options. This makes it possible to identify the treatment with the best chance of success.
Using the information provided by this analysis, the doctor and patient can choose a personalised care path together. The aim is to increase the patient's quality of life.
Personalised care, shorter care pathways and reduced workload
The approach should simultaneously reduce pressure on oncology. The number of patients within the healthcare sector as a whole and specifically oncology is expected to increase explosively over the next decade. Currently, patients go through all the steps of an oncology pathway by default. Therefore, in practice, the pathway is the same for every patient.
The deployment of advanced AI models and systems should change this. Healthcare professionals can use these to accurately determine which steps are or are not necessary for a specific patient. Not only does this relieve care, it also allows them to offer personalised care and shortened care pathways. This should make care more effective and efficient, benefiting both patients and doctors.
Complete picture of medical history often missing
However, embracing AI in healthcare is not easy. A key challenge is that many current models are static and integration with the electric patient record is lacking in many cases. Consequently, healthcare organisations often do not have a complete picture of a patient's medical history.
Using new AI models, Info Support and the NWO aim to solve this. The models will support both patients and doctors in developing personalised oncology pathways.
Explainability of AI models
However, the collaboration is broader. A good part of the research is focused on the explainability of AI models. An important topic, as currently AI functions to some extent as a kind of 'black box'. How and why a model draws a particular conclusion is not always clear. This makes interpreting conclusions from AI systems and relying on them in practice difficult.
Info Support and the NWO therefore want to make AI models more explainable and transparent. In this way, healthcare should be able to trust the conclusions drawn by algorithms earlier and with more certainty.
Machine learning platform
Within the collaboration, Info Support will provide access to a machine learning platform, among other things. The NWO will use this to train and develop reproducible models. In addition, Info Support will provide training on Explainable AI, MLOps and Advanced Python Programming within Microsoft Azure environments. The company will also provide an AI expert to advise the NWO and support the NWO in mentoring postdoc researchers.
Joop Snijder, Head of Research Center AI at Info Support: "We are extremely proud to contribute to research to make oncology care more personal and efficient. We find it important to engage in corporate social responsibility. With our Research Centre, we want to make a positive impact on fundamental and societal issues at stake. Through our efforts, Info Support hopes to do its bit to keep healthcare accessible and ease the workload."
Author: Wouter Hoeffnagel
Image: Gordon Johnson via Pixabay
