Session Report: Acute stroke imaging – core, perfusion and beyond – Joint Session with the Society of Vascular and Interventional Neurology (SVIN)

This joint session with the Society of Vascular and Interventional Neurology was chaired by Götz Thomalla (Hamburg, Germany) and Thanh N. Nguyen (Boston, United States), bringing together leading experts to explore the evolving role of imaging in acute stroke, from classical concepts such as core and penumbra to emerging technologies and future perspectives.

Leptomeningeal collaterals: from physiology to therapeutic target

Marilyn Cipolla (Burlington, USA) opened the session by exploring the critical role of leptomeningeal collaterals in acute ischemic stroke. These vessels provide low but sufficient perfusion to sustain penumbral tissue viability until reperfusion can be achieved. Preclinical models demonstrated that induced hypertension could enhance collateral flow and reduce infarct core size. While randomized trials have shown early neurological improvement, long-term benefits remain uncertain.

Other promising strategies aimed at improving collaterals include:

• Sphenopalatine ganglion stimulation, showing modest increases in cerebral blood flow, particularly in cortical strokes;
• Remote ischemic conditioning, effective in experimental models but not confirmed in large preclinical trials;
• Head-down tilt positioning, associated with early neurological improvement and reduced infarct growth in experimental settings.

Overall, collaterals represent a promising avenue to extend the therapeutic window, although patient selection and optimal strategies remain to be defined.

Is the core–penumbra concept still relevant?

Tudor Jovin (Camden, USA) revisited the cornerstone concepts of ischemic core and penumbra in acute stroke. While these concepts have historically guided treatment decisions, especially in extended time windows, he argued that current imaging techniques inadequately capture the biological complexity of the ischemic tissue. Evidence from large core thrombectomy trials demonstrated substantial treatment benefits even in patients with minimal or no apparent mismatch.

Key insights include:

• No clear treatment effect modification based on mismatch status in thrombectomy studies;
• Similar treatment effect sizes in large-core and small-core patients;
• Imaging “core” can overestimate irreversible injury due to phenomena such as lesion reversal and “ghost infarcts”.

Importantly, penumbral salvage explains only a fraction of clinical benefit. Other mechanisms, such as edema reduction and preservation of the neurovascular unit, play a major role. He concluded that while the concept remains biologically relevant, it should be considered only one of several mechanisms underlying clinical benefit with reperfusion.

Photon-counting CT: a new era in stroke imaging

Linda Jacobi-Postma (Maastricht, Netherlands) presented photon-counting CT as a transformative innovation in neuroimaging. Compared to conventional CT, this technology offers: Higher spatial resolution, reduced electronic noise, improved contrast-to-noise ratio, lower radiation and contrast dose, intrinsic spectral imaging capabilities.

In stroke imaging, photon-counting CT improves:

• Detection of early ischemic changes and small infarcts;
• Visualization of posterior fossa and brainstem lesions;
• Identification of distal and medium vessel occlusions;
• Characterization of plaque composition and vessel pathology.

It also enables advanced applications of photon-counting CT such as material decomposition (e.g., distinguishing hemorrhage from iodine) and improved post-interventional imaging.

Despite its promising characteristics, optimization of acquisition parameters and clinical integration remain ongoing challenges. Nonetheless, this technology may represent not just an incremental improvement, but a paradigm shift in CT imaging.

Artificial intelligence in acute stroke imaging: from promise to proof

Sunil Sheth (Houston, USA) addressed the rapidly expanding role of artificial intelligence (AI) in stroke workflows. AI-based tools are now widely implemented for large vessel occlusion (LVO) detection, automated imaging analysis, workflow coordination and communication. However, he emphasized a critical point: performance accuracy alone is not sufficient. AI must demonstrate clinically meaningful impact. Using a stepped-wedge study design across multiple stroke centers, his group showed that AI implementation:

• Reduced workflow times (e.g., door-to-reperfusion metrics);
• Improved efficiency, particularly when actively used by clinicians.

Importantly, benefits were linked to user engagement, highlighting that technology alone is insufficient without proper integration into clinical practice. Looking ahead, AI may move beyond detection tasks toward predictive modeling, helping clinicians anticipate outcomes such as hemorrhage risk or long-term recovery.

Imaging in haemorrhagic stroke: current standards and future directions

Martina Goeldlin (Bern, Switzerland) focused on acute imaging in intracerebral hemorrhage (ICH), which accounts for approximately 15% of strokes. She emphasized that: rapid imaging is equally critical in ICH (“time is brain”) and CT remains the primary modality, though MRI can be used in selected centers

Key challenges include:

• Identifying underlying etiology (e.g., vascular malformations, small vessel disease, cerebral amyloid angiopathy);
• Avoiding delays in treatment caused by ischemic stroke–oriented workflows (e.g., direct-to-angio approaches).

Emerging imaging markers and scores, such as those for cerebral amyloid angiopathy, may improve risk stratification and guide management. Advanced imaging techniques and better integration of MRI may further refine diagnosis and prognosis in hemorrhagic stroke.

Take-home message

This session underscored a major shift in stroke imaging:

• Imaging is evolving from a diagnostic tool to a therapeutic guide;
• Traditional concepts like core–penumbra are being redefined in light of new evidence;
• Technologies such as photon-counting CT and AI are reshaping clinical workflows.

A more individualized, biology-driven approach to stroke care is emerging

As innovation accelerates, the challenge will be not only adopting new tools but rigorously validating their impact on patient outcomes.