Background and purpose: A significant amount of strokes are reported to be diffusion-weighted imaging (DWI) negative in acute imaging. We attempted to quantify the rate of false-negative high-resolution (hr) DWI and to identify a valid screening tool to guide follow-up MRI to diagnose infarction initially not visible on hrDWI.
Methods: An a priori-defined post hoc analysis of a prospective 3T MRI cohort of acute cerebrovascular events imaged within 24 hours of ictus. Basic demographics, risk factors, National Institute of Health Stroke Scale, and imaging parameters were recorded.
Results: Of 151 patients with negative acute hrDWI, 63 received follow-up scans depicting infarction in 7 cases (11.1%). Persistence of clinical symptoms as established by National Institute of Health Stroke Scale on the following day was strongly associated with infarction on follow-up MRI (odds ratios, 17.5; 95% confidence interval, 2.83-108.12). Negative predictive value of follow-up National Institute of Health Stroke Scale was 0.96.
Conclusions: Infarcts are frequently invisible on initial hrDWI, but we may well trust in negative hrDWI in completely transient cerebrovascular events.
Keywords: diffusion-weighted MRI; infarction; transient ischemic attack.