Mobiele menu

Using a big data approach to divide the syndrome Rheumatoid Arthritis into homogenous subsets


Klinische patroonherkenning 2.0

Het menselijk brein schiet tekort om de patronen binnen reumatoïde artritis te herkennen. Met toepassing van artificiële intelligentie op DNA, bloed en elektronische dossiers van 10,000 patiënten gaan onderzoekers de (sub)ziektes van reumatoïde artritis identificeren. Deze resultaten kunnen de behandeling en het onderzoek verbeteren naar de oorzaken van rheumatoïd2 artritis verbeteren.

Clinical pattern recognition 2.0

The human brain fails to recognize the patterns within rheumatoid arthritis. By applying artificial intelligence on data from DNA, blood and electronic records of 10,000 patients, researchers will identify the (sub)diseases of rheumatoid arthritis. These results can improve treatment and research into the causes of rheumatoid arthritis.


Samenvatting van de aanvraag

Rheumatoid arthritis (RA) is a syndrome characterized by inflammation and subsequent destruction of the joints. It leads to chronic pain, disability, work loss, social isolation and additional comorbidities such as cardiovascular diseases. This autoimmune syndrome develops years before the onset of symptoms. The etiologic factors are largely unknown which makes disease classification and treatment imprecise. The group of patients that are classified as RA varies in treatment response, long-term outcomes and comorbidities. Clearly, the heterogeneous syndrome called RA is a collection of different more homogeneous diseases. The aim of my project is: A. to identify the underlying homogeneous (sub)diseases in the heterogeneous group of RA patients, thus improving the taxonomy of subsets of RA; B. to identify the health events preceding the development of RA (sub)diseases; and C. to link the clusters with treatment response thereby identifying the optimal treatment regimes for the different RA (sub)diseases. To achieve these aims, I will embark on innovative high-throughput analytic approaches to identify patterns of similarity in the high dimensional clinical data. I will use in depth, combined analyses of the different pillars of heterogeneity: genetics, medical history, serology and clinical symptoms. Several (international) cohorts and subsequent big data sets are available through the international projects on which I collaborate. I will collect longitudinal EMR data from ten years prior to disease onset from hospitals and primary care as well as serology and genetics data. There are five pillars of heterogeneity among the patients with RA: genetics, medical history, serology, clinical symptom and, treatment response. I will create analytical pipelines that make optimal use of the information captured by these different pillars as well as of the time relationship between the factors. This project will serve both patients and research. The improved taxonomy of RA will support clinical decision making and improve personalized treatment. Additionally, the subset specific features will better define distinct RA endotypes and guide further fundamental research. Simultaneously, the homogenous subsets will facilitate the development of targeted drugs. Being trained as rheumatologist and having completed a computational post-doc at Harvard, my scientific profiles as bridge builder neatly fits the requirements to accomplish this project successfully.



Looptijd: 52%
Looptijd: 52 %
Gerelateerde subsidieronde:
Projectleider en penvoerder:
dr. R Knevel
Verantwoordelijke organisatie:
Leids Universitair Medisch Centrum