David Marlevi

David Marlevi

Biträdande Lektor | Docent
E-postadress: david.marlevi@ki.se
Besöksadress: Eugeniavägen 3, Karolinska universitetssjukhuset, NKS A8:01, 17176 Stockholm
Postadress: K1 Molekylär medicin och kirurgi, K1 MMK Klinisk fysiologi, 171 76 Stockholm
Del av:

Om mig

  • 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
    "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
    system.

     
    * 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,
    2014
    * Endeavour Research Award, Australian Government Research Award fellowship
    * Henrik Göransson Sandviken scholarship, 2011
    * Hjalmar Berwalds minne för framstående matematiska studier, 2010

Forskningsbeskrivning

  • /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.

Artiklar

Alla övriga publikationer

Forskningsbidrag

  • Europeiska forskningsrådet
    1 May 2023 - 1 May 2028
  • Swedish Research Council
    1 January 2023 - 31 December 2026
    What will be done? The aim of this proposal is to provide a framework for non-invasive, image-based assessment of pressure gradients throughout the cardiovascular system, where critical limitations of spatial, temporal, and flow scales are overcome in a unique multiscale approach.How will it be done? Combining full-field magnetic resonance imaging (4D Flow MRI) with physics-informed, neural network-based image analysis, a set of specific aims will allow for pressure gradient assessment across spatial (big/small vessels), temporal (fast/slow flows), and flow (laminar/turbulent flows) scales. Being trained and validated across dedicated in-silico, in-vitro, and in-vivo cohorts, the framework will be ready for direct clinical use, enabling imaging of pressure gradients across previously inaccessible cardiovascular scales (heart/aorta/brain).Why is this important? Estimation of cardiovascular pressure gradients is critical for diagnosis and treatment of many cardiovascular diseases. Still, for a large number of conditions, routine assessment is hindered by complex 3D flow, poor spatiotemporal image resolution, or turbulence-driven flow.Why us? The PI is a recognised expert in quantitative translational flow imaging. Now returning from a fellowship at the Massachusetts Institute of Technology, this 4-year project will establish this unique line of research in Sweden, gathering a dedicated set of collaborators from technical and clinical fields to ensure feasibility and impact.

Anställningar

  • Biträdande Lektor, Molekylär medicin och kirurgi, Karolinska Institutet, 2025-2030
  • Postdoktor, Molekylär medicin och kirurgi, Karolinska Institutet, 2021-2024
  • Postdoctoral Researcher, Institute for Medical Engineering & Science, Massachusetts Institute of Technology, 2019-2021

Examina och utbildning

  • Docent, Medicinsk bildvetenskap, Karolinska Institutet, 2025
  • MEDICINE DOKTORSEXAMEN, Institutionen för kliniska vetenskaper, Danderyds sjukhus, Karolinska Institutet, 2019

Nyheter från KI

Kalenderhändelser från KI