David Marlevi

David Marlevi

Postdoctoral Researcher
Visiting address: Eugeniavägen 3, Karolinska universitetssjukhuset, NKS A8:01, 17176 Stockholm
Postal address: K1 Molekylär medicin och kirurgi, K1 MMK Klinisk fysiologi, 171 76 Stockholm

About me

  • I am a research lead in quantitative cardiovascular imaging, developing
    data-driven image analysis tools to tackle urgent clinical challenges across
    the heart, aorta, and brain.
    I am a biomedical engineer with a focus on translational cardiovascular
    imaging. Specifically, I am intrigued by the translation of image-based
    engineering into clinical practice, using data-driven utilities to enhance
    diagnosis, improve prognosis, and provide fundamental mecanistic
    understanding of cardiovascular disease.
    After graduating from the joint doctoral program in Medical Technology from
    the Royal Institute of Technology (KTH) and KI with a thesis entitled

  • dswid=7561">Non-invasive imaging for improved cardiovascular care", I spent two 

  • years as a postdoctoral fellow at the Massachusetts Institute of Technology
    (MIT), funded by a Knut and Alice Wallenberg foundation scholarship and
    working under the tutelage of Prof. Elazer R. Edelman (edelmanlab). At 

  • MIT, I worked on AI-driven image analysis to monitor intravascular
    interventions, as well as lead lab efforts on novel vascular drug delivery
    systems. In 2021, I returned to Sweden and KI as a research lead in
    quantitative cardiovascular imaging, working closely with clinical and
    technical fellows at both KI and the Karolinska University Hospital to
    translate advanced image technologies into clinical practice. Specific
    focuses has been on hemodynamic mapping by full-field phase-contrast magnetic
    resonance imaging (4D Flow MRI), with coupled physics-informed analysis
    allowing for regional hemodynamic quantifications across the cardiovascular


  • * Early Career Award – Translational Science, Society for Cardiovascular
    Magnetic Resonance (SCMR), 2021
    * Runner-up, Best presentation in Basic science, 1st Annual Marvin M. Kirsh
    Resident Research Symposium, University of Michigan, 2021
    * Potchen-Pasariello Award – Best presentation in Clinical Science,
    Society for Magnetic Resonance Angiography (SMRA), 2020
    * Trainee grant, IEEE Nuclear Science Symposium and Medical Imaging
    Conference, 2015
    * Travel award, IEEE International Ultrasonics Symposium, 2015
    * KTH Best graduate student of the year, KTH Royal Institute of Technology,
    * Endeavour Research Award, Australian Government Research Award fellowship
    * Henrik Göransson Sandviken scholarship, 2011
    * Hjalmar Berwalds minne för framstående matematiska studier, 2010


  • Non-invasive estimation of cardiovascular pressure gradients:
    Regional quantification of cardiovascular pressure gradients is critical for
    diagnosis, treatment planning, and risk prediction of many cardiovascular
    disease. Still, for a large number of conditions, non-invasive assessment is
    obstructed by inherent method limitations, and a wide range of clinical
    instances exist where regional pressure behaviour remains unexplored. To
    tackle this, we have recently deployed a combination
    of physics-informed image analysis (invoking fundamental fluid mechanical
    description of blood flow) and full-field flow imaging (4D Flow MRI) to allow
    for arbitrary probing of pressure gradients across previously inaccessible
    compartments. Here, we seek to extend these utilities to further understand
    early hemodynamic changes indicative of later physiological impairement,
    including validation, implementation, and clincial utility across spatial
    (large / small vessels), temporal (fast / slow flows) and flow (laminar /
    turbulent) scales.


  • Super-resolution 4D Flow MRI:
    The advent of full-field flow imaging by 4D Flow MRI has fundamentally
    changed our ability to interrogate complex hemodynamic behaviour in a direct
    clinical setting. However, spatiotemporal limitations exist based on the
    clinical time frames in which the systems can be used, obstructing assessment
    of regional or highly transient flow events. To tackle this, we have recently
    employed deep residual networks to enhance spatial image resolution,
    effectively pushing quantitative 4D imaging into challenging intracranial
    vessels. Now, we seek to extend the same utilities into temporally
    challenging flows such as in the heart, or through complex aortic disease.
    Further inclusion of so called physics-informed networks are also expected to
    expand clinical impact and versatility of 4D Flow MRI across a wide
    cardiovascular application range.


All other publications


  • Postdoctoral Researcher, Department of Molecular Medicine and Surgery, Karolinska Institutet, 2021-2024
  • Postdoctoral Researcher, Institute for Medical Engineering & Science, Massachusetts Institute of Technology, 2019-2021

Degrees and Education

  • Doctor Of Philosophy, Department of Clinical Sciences, Danderyd Hospital, Karolinska Institutet, 2019

News from KI

Events from KI