Overview of Neuroimaging of Stroke

Introduction

Stroke incurs ischemic or hemorrhagic injury in the brain, typically associated with alterations in the cerebral circulation, including a variety of disorders that affect arteries, capillaries, and venous components. Imaging provides detailed information on brain tissue and vascular anatomy for clinicians to corroborate their clinical history and examination findings in acute stroke cases. The significance of imaging in stroke depends on unraveling the pathophysiology of acute ischemia and plays a pivotal role during the acute phase to formalize therapeutic plans. This chapter provides an overview on the role of neuroimaging in stroke, including computed tomography (CT), magnetic resonance imaging (MRI), and cerebral angiography with further elaboration in subsequent chapters. Across a variety of practice scenarios, clinicians often have a choice of various imaging modalities to yield further information on the care of patients with stroke. Imaging data serve as an extension of the clinical examination that routinely enhances the clinical evaluation and neurological localization in a patient with stroke. The following discussion delineates the key goals of neuroimaging in stroke to augment clinical decision making.

Overwhelmingly, the primary goal of imaging using noncontrast CT or MRI has been to differentiate hemorrhage from ischemia for consideration of intravenous (IV) thrombolysis. There has been a trend toward identifying early ischemic changes along with attendant early vessel signs on CT or MRI. “Time is brain” has been the mantra for decades, whereas recent advanced imaging modalities have enticed clinicians to modify it to “tissue is brain,” largely disclosed by imaging. It has been shown that time alone is rudimentary and neuroimaging complements decision making in acute stroke cases.

In recent years, more comprehensive imaging techniques such as multimodal CT or MRI that encompasses angiography and perfusion images (further details in following subsections), have emerged that provide a detailed perspective of cerebrovascular disease pathogenesis, vascular anatomy, and the functional correlates of perfusion abnormalities. These imaging strategies have been used on a regular basis as they portray a snapshot of brain parenchyma and guide further therapeutic decisions. Various key elements of pathophysiology portrayed with imaging include large vessel occlusion (LVO), collateral blood flow patterns, resultant hemodynamic modification in blood flow, and the tissue injury that ensues.

Commonly Used Imaging Modalities

CT comprised the predominant imaging modality to assess patients with stroke from the acute to chronic phases in previous eras as compared with the vast neuroimaging data available via multimodal CT/MRI more recently. Since the advent of MRI and concomitant newly refined multimodal techniques of vascular neuroimaging, the understanding of stroke has expanded in numerous directions. For acute stroke cases, multimodal CT typically includes noncontrast CT, CT angiography of head and neck (CTA), as well as CT perfusion (CTP). Multimodal MRI includes various sequences, such as diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC), fluid-attenuated inversion recovery (FLAIR), gradient recalled echo (GRE), and perfusion-weighted imaging in addition to MR angiography of head and neck (MRA) ( Fig. 130.1 ).

CT has been the standard for initial and subsequent comparisons of imaging lesion evolution, yet limitations include failure to reliably discern lesions that are very early from stroke symptom onset and those located in specific locations such as the posterior fossa . MRI with its numerous imaging sequences including DWI provides readily available data to prioritize and make decisions for thrombolysis or endovascular therapy, if applicable. The decision between CT and MRI is often based on an individual case approach or the availability of either technique in each stroke center. CT is the optimal imaging modality for MRI-incompatible patients due to specific pacemaker devices and implantable metallic objects or those patients who are hemodynamically unstable to remain supine for an extended period of time. Image acquisition using MRI may impose greater logistical challenges compared with CT, although it provides greater details and dimensions of brain tissue and corresponding vascular changes. Various institutions have developed multimodal CT or MRI protocols with demonstrated feasibility and utility in the early management of patients with stroke across acute settings.

Ischemic Stroke

Ischemic stroke is a dynamic process that evolves from acute through subacute and chronic phases. Imaging in stroke focuses to determine the diagnosis and etiology, lesion localization, extent of ischemic evolution, therapeutic implications, and expected prognosis. Imaging evaluation in acute ischemic stroke (AIS) often revolves around the key parameters of core infarct volume, presence and extent of perfusion-diffusion mismatch, and collateral flow profiles or grades that refine therapeutic decisions. CT provides a solid platform for physicians to not only exclude hemorrhage, but also estimate the extent of infarct evolution. Rather than focusing solely on ruling out massive infarcts, the trend has shifted to evaluate subtle signs of ischemia including hypoattenuation or obscuration of the lentiform nuclei, loss of the insular ribbon, sulcal effacement, cortical hypodensity, and the presence of various hyperdense vessel findings. Alberta Stroke Program Early CT score (ASPECTS) is a topographic scoring system to assess early ischemic parenchymal changes on initial CT in patients with AIS of anterior circulation. Interobserver discrepancy has emerged as a major concern even among experienced stroke neurologists, to interpret early signs of ischemia using initial CT scans .

CT has been the primary neuroimaging modality with rapid and widespread availability, although increasing number of institutions are resorting to MRI-based technique that confers extensive data on cerebrovascular events. MRI involves multiple sequences that provide snapshots of brain tissue including prior silent ischemia and other cerebrovascular pathology, in exquisite detail. DWI is the primary sequence used in patients with AIS that reveals restricted Brownian motion of water molecules secondary to cytotoxic edema. DWI has a sensitivity of 99% and a specificity of 92% in detecting ischemic changes as compared with CT . DWI patterns of ischemia are especially pathognomonic for disease mechanisms pertaining to LVO, thromboembolism, small-vessel disease, hypoperfusion or border zone mechanisms ( Fig. 130.2 ). FLAIR sequences on MRI may reveal hypointense signal for space-occupying lesions with surrounding vasogenic edema, whereas old infarcts are hyperintense due to increased water content in brain tissue. Periventricular FLAIR hyperintensities portray old small-vessel lacunar infarcts and further guide clinicians in formulating therapeutic plans for risk management of their patients. FLAIR vascular hyperintensities may also offer subtle clues about the presence of slow flow in collateral vessels downstream of an arterial occlusion.

Therapeutic decision making in AIS primarily involves DWI/ADC and FLAIR sequences once hemorrhage has been excluded on GRE. DWI hyperintensities with ADC hypointensity indicate cytotoxic edema associated with acute to subacute infarcts. Furthermore, lesions with mismatch between DWI and FLAIR sequences depict temporal aspects and the duration of ischemia . Neuroimaging usually compliments the examination findings obtained by clinicians, whereas lesion patterns are particularly helpful in stroke. Distally located cortical lesions likely may infer thromboembolic phenomena. The proximal nidus or embolic source may be artery–artery or from the heart, prompting further diagnostic testing. Subcortical, especially basal ganglia and deep white matter, lesions are more typically due to microvascular in situ disease or small-vessel ischemia, although proximal subocclusive lesions involving lenticulostriate branches of the middle cerebral artery (MCA) can also produce similar lesions. Border zone lesions of ischemia between the principal arterial territories in the brain may manifest as a string or archipelago of discrete cortical and subcortical lesions that implicate a proximal stenotic or occlusive lesion causing hypoperfusion.

Contrast enhancement of ischemic lesions is helpful in the age determination of any vascular lesion. Parenchymal enhancement is associated with disruption of the blood–brain barrier that usually presents in a gyriform or ring-like shape, located peripheral to the central ischemic lesion. It usually becomes visible around 4–7 days after the ischemic event, tends to resolve within 8 weeks, and may persist beyond 3 months in isolated cases. Serial imaging of infarct evolution over ensuing days or weeks illustrates the dynamic nature involved in cerebral ischemia and various stages may be encountered, including “fogging” or the transient disappearance of subacute lesions before chronic scar formation. Fogging may be seen with either CT or MRI and refers to gradual fading of lesion on follow-up serial scans usually 2 weeks after the onset of ischemia.