Elham Rostami

Elham Rostami

Researcher
Visiting address: Solnavägen 9, Biomedicum C5, 17165 Solna
Postal address: C3 Fysiologi och farmakologi, C3 FyFa Neurotrauma och translational neurokirurgi, 171 77 Stockholm
Part of:

About me

  • I completed my medical degree (MD), internship (AT) and PhD at Karolinska Institutet, in the field of neurotrauma and translational neuroscience. After my PhD, I moved to Uppsala University Hospital for a research residency in neurosurgery. I became a board‑certified neurosurgeon in Sweden in 2018 and was appointed Associate Professor (docent) in 2017. Since then, I have developed an independent research programme in traumatic brain injury and brain resilience, while maintaining a broad clinical neurosurgical practice and building national and international collaborations that link Karolinska Institutet, Uppsala University and global neurotrauma networks.

    I am currently Chair of the European Association of Neurosurgical societies (EANS) Section of Trauma Neurosurgery (since 2025) and Vice President of the Swedish Young Academy, where I work on research policy, academic conditions for early‑career researchers and international collaboration in neurotrauma.”

    I Was also recently appointed as Director of HumanLab at KI.

Research

  • My research interests centre on traumatic brain injury (TBI), brain resilience and precision medicin in neurotrauma and neurosurgery. I am particularly interested in why patients with apparently similar injuries follow very different recovery trajectories, and how host biology, acute pathophysiology and life‑course factors interact to shape outcome.

    A major focus is on host‑specific determinants of TBI outcome, including genetics, polygenic risk, and multi‑omic signatures, which I study through international cohorts and population biobanks integrated in the newly formed GEN‑TBI‑BRAIN consortium. In parallel, I work with inflammatory and neurochemical markers, neuromonitoring data and advanced imaging to characterise acute brain responses and link them to long‑term function. Swedish nationwide registers and other large‑scale data sources are central tools to understand long‑term psychiatric, neurological, somatic and socioeconomic consequences after TBI.

    Another key interest is the development and validation of AI‑based prognostic models and decision‑support systems that integrate clinical, imaging, neuromonitoring and biological data. My goal is to move from descriptive prognostication towards mechanism‑based stratification and personalised interventions in neurotrauma and neurosurgery. On the experimental side, I have been working with different types of experimental TBI models developed in Professor Mårten Rislings lab.  Currently I collaboret with Carl Sellgren on using a human‑relevant preclinical platforms, including humanised cerebral organoid TBI models, to test hypotheses and candidate therapies that emerge from clinical and omics studies. 

    Finally, I am increasingly interested in brain resilience beyond classical TBI, for example in operational and extreme environments such as aviation and hypergravity exposure. Through HumanLab and ExtremeLab at Karolinska Institutet, I aim to build holistic multimodal human studies that combine physiology, imaging, cognition, biosampling and digital behaviour to understand brain and performance resilience under stress and to inform safer, more personalised neurosurgical and neurocritical care.

Articles

All other publications

Grants

  • Swedish Research Council
    1 January 2025 - 31 December 2028
    Traumatic Brain Injury (TBI) affects over 60 million people annually worldwide, leading to significant disabilities and increasing the risk of dementia. Despite TBI´s heterogeneity and varying outcomes among similarly injured patients, current treatment relies on a uniform, one-size-fits-all approach, often resulting in suboptimal care. This one-size-fits-all approach fails to accommodate the individual variations and complexities inherent in each case of TBI, leading to suboptimal outcomes. Recognizing the need for personalized treatment strategies, this project proposes the integration of clinical variables, head CT-scans, biological, high-resolution physiological data, and genetic variations, cognitive reserve, harnessing comprehensive Swedish and international TBI datasets, bolstered by extensive Swedish linkage registers. Employing machine learning algorithms, this approach strives for customized interventions and accurate prognosis tools, transforming care in neurointensive clinics. Additionally, addressing the gap between animal studies and human trials, we introduce a humanized in vivo model for TBI research, essential for testing neuroprotective drugs and understanding human cell responses. Combining clinical and experimental efforts with advanced data-driven techniques, our project targets the discovery of effective drug targets to improve TBI recovery, underscored by a humanized model for direct relevance to patient care.
  • Swedish Research Council
    1 January 2024 - 31 December 2027
    This proposal builds on multi-disciplinary data describing abnormal synaptic pruning as a mechanism in schizophrenia (SZ). So far, increased expression of the SZ risk gene coding for complement component 4A (C4A) has been shown to contribute to excessive microglial engulfment of synaptic structures in patient-derived SZ models, while protein levels of C4A are elevated in first-episode SZ patients. However, the mechanistic understanding of the molecular events that leads to activity-dependent synapse elimination by microglia, and other glia cells, is largely incomplete due to the lack of adequate experimental models. This has complicated the identification of suitable drug targets and stalled the highly needed clinical studies using novel or repurposed compounds in SZ. Here, we propose a novel and unique integrated experimental and clinical approach - linked on subject level and supported by humanized in vivo models - to identify suitable drug targets for modifying glial synapse elimination in SZ. Using state-of-the-art organoid models combined with high-resolution single cell transcriptomic profiling we aim to identify cell types and mechanisms contributing to excessive pruning in SZ. By perturbing these mechanisms in our models, we select a set of candidate mechanisms that are validated in vivo and in a clincial context (CSF and longitudinal MRI imaging) using within-subjects analyses followed by case-control analyses utilizing a cohort consisting of &gt
    200 subjects.

Employments

  • Researcher, Department of Physiology and Pharmacology, Karolinska Institutet, 2026-
  • Researcher, Department of Neuroscience, Karolinska Institutet, 2021-2025

Degrees and Education

  • Degree Of Doctor Of Philosophy, Department of neuroscience, Karolinska Institutet, 2012
  • University Medical Degree, Karolinska Institutet, 2008

News from KI

Events from KI