The aim of this systematic review is to perform a literature search to identify and analyze key publications that have incorporated neuroimaging modalities in stroke patients to predict their clinical outcomes. 

A scoping review was conducted using PubMed, Embase, Scopus, Web of Science, and Cochrane databases. Studies published in English from 2000 to 2023 were screened. Inclusion criteria encompassed original research, meta-analyses, and systematic reviews evaluating neuroimaging for prognostication in ischemic stroke. Case reports, editorials, and non-human studies were excluded. Relevant findings were synthesized and grouped by imaging modality.

Conventional CT remains the cornerstone for initial assessment. Low ASPECTS and presence of hyperdense MCA sign were associated with poor outcomes. On CT angiography, poor collateral status, larger clot burden, and clot length >8 mm were linked to poor outcomes. HERMES collaborators demonstrated that large initial core volume on CT perfusion was associated with poor outcomes. MRI final infarct volume >22 ml predicted poor 90-day clinical outcomes. Loss of corticospinal tract diffusivity in MRI diffusion tensor imaging during acute stroke was associated with poor outcomes, while improvement in fractional anisotropy correlated with motor recovery. Functional MRI provided insights into network reorganization and motor recovery potential.

Neuroimaging is central to outcome prediction in acute ischemic stroke. Each modality contributes unique prognostic information, and their combined use enables comprehensive assessment. Further research should focus on incorporating multimodal imaging techniques and exploring machine learning algorithms to develop improved stroke outcome prediction models.