One advantage of MRI is that there are numerous types of contrast (e.g., T1, T2, diffusion, spectroscopy, etc.) that are available to assist in the diagnosis of SAH. Standard T1- and T2-weighted images typically have very poor SAH contrast (7); however, other MR sequences can be very sensitive and specific for SAH (8). Proton density MRI was found to have similar findings in patients to that on CT at early time points (<6 or <12 h post-SAH). T1 was also reported to change but with less accuracy and only at >12 h (9, 10). Traditional T2-weighted imaging has also been useful for the evaluation of SAH but primarily within the context of monitoring edema, although if the region of extravasated blood is sufficiently large, T2 can visualize these changes.

T2*-weighted imaging (T2*WI) can be reliably used to identify regions containing blood products. T2* images are also very sensitive to SAH, having 94% sensitivity in the acute stage (<4 days) and 100% sensitivity in the subacute sage (>4 days). This is comparable to CT in the acute, and superior in the subacute periods (7). The iron in deoxygenated hemoglobin and blood breakdown products cause strong T2* shortening which gives excellent SAH contrast. However, T2* images are susceptible to artifacts near air–tissue interfaces and may obscure subarachnoid regions and prevent detection of SAH.

Diffusion-weighted imaging (DWI) is often used in stroke as the neuroimaging modality of choice due to its sensitivity to ongoing tissue injury that is reflected by decreased proton diffusion. However, DWI and its quantitative parameter, the apparent diffusion coefficient (ADC), have not been routinely used in clinical or basic research with regard to SAH. However, several clinical reports have reported that use of DWI (or ADC) is useful for identifying ischemic tissue regions as a result of the SAH (9, 11–15). In virtually all of these clinical reports, the authors demonstrate that the DWI accurately reports the ischemic injury near the location of the SAH injury or in proximity to vessels where the SAH injury occurred.

Weidauer et al. (12) reported cortical banding infarcts after SAH that were visualized by DWI. Similarly, Sato et al. (13) demonstrated that aggressive clinical management of patients that showed DWI or ADC abnormalities, irrespective of the relative ADC (ratios) change, resulted in better outcomes for the patients. Other studies have also reported the value of ADC in SAH. Liu et al. (16) showed that ADC, but not T2-weighted imaging nor DWI, could detect broad changes in frontal lobe white matter that resolved over the course of 1 year.

Animal studies of SAH and DWI have been limited. Moseley and colleagues (17, 18) in a series of reports detailed the use of DWI to monitor the very early (within minutes) DWI and ADC changes following induction of SAH. They reported that shortly after SAH induction there was a pattern of diffusional changes within the brain that appeared to reflect the initial acute vasospasm that then appeared to initiate spreading depression, a wave of electrical depolarization, within the brain tissues. Others have reported use of DWI and have showed changes on DWI but no significant conclusions were reached (19).

Fluid-attenuated inversion recovery imaging (FLAIR) is an excellent MR sequence as it suppresses all fluid signals within the brain allowing improved visualization of SAH in the absence of cerebrospinal fluid (CSF) or edema (Figs. 1 and 2). FLAIR is a T2-weighted sequence that has an inversion recovery pulse added to null the signal from CSF, which is normally very bright on T2 images. SAH causes a hyperintense signal on T2-weighted images. In a standard T2 image this hyperintensity would be hidden by the bright CSF, thus, nulling the signal from CSF allows SAH to be clearly visualized.

FLAIR was the first MRI sequence to reliably detect acute SAH (20). Since then FLAIR has been shown to detect SAH in many studies, particularly those comparing it to CT. In clinical settings, it has been shown to be at least as sensitive as CT (10). However, FLAIR and T2* MRI has been reported to overestimate the SAH or ICH volume while CT underestimates the lesion volume (21). In vitro models have shown it more sensitive than CT (22), and some animal models have shown FLAIR to be clearly better than CT as CT was inadequate to detect SAH in the reported model (23). However, in animal studies some caution is required in comparison to human data due to the type of SAH or ICH model that may change the visibility of lesions. An example is the study by Küker et al. who were able to visualize the ICH and SAH lesions at both 1.5 and 0.5 T on T2* and FLAIR with the authors reporting that FLAIR was optimal for SAH. In this study, the model used venous blood to induce the injury which could account for the improved visualization on MRI at the hyperacute time points (24).

Other MRI modalities have been reported but are not widely used either clinically or in basic research. These include the use of perfusion-weighted imaging (PWI), magnetic resonance spectroscopy (MRS), and derived parameters including relative blood volume maps. Spectroscopy for SAH injury markers from CSF of patients has also been performed but these are typically not used routinely in the clinic (3). A recent report of 31P MRS in patients receiving hypermagnesium therapy for SAH reported an effective therapeutic potential (25).

More recently, the introduction of susceptibility-weighted imaging (SWI) has vastly improved the ability to visualize extravascular blood deposition within the brain (Figs. 1 and 2). The technical and clinical applications relevant to adults have been recently reviewed by Haacke and colleagues (26, 27). SWI applications to pediatric cases have also been reviewed (28–30). While SWI is uniquely sensitive to blood and blood products within the brain, it has not been routinely used for SAH diagnosis nor clinical evaluation. One of the potential reasons for this is that the complementary phase information that is essential for the dramatic SWI contrast is problematic at air tissue interfaces, thus potentially making SAH diagnosis more difficult, particularly at the base of the skull. However, advances in phase mapping (31) and phase unwrapping (32) may alleviate this issues.

A recent publication evaluating traumatic SAH using SWI demonstrates the utility of this advanced technique for diagnosis (33). In 20 patients, the authors evaluated eight anatomical regions within the subarachnoid space and reported that SWI was able to identify five additional patients with SAH that could not be diagnosed with conventional CT.

Animal studies have reported increased sensitivity to blood products in experimental models (34, 35) (Fig. 3). In our own studies, we have shown that SWI, even using only the magnitude images, can delineate the sequence of events as extravascular blood is degraded within the brain (34). Since SWI is a new technique that requires additional filtering and postprocessing that are largely not yet available commercially from imaging manufacturers, we delineate a detailed protocol for acquisition, processing, and analysis of SWI-derived SAH data. Unequivocal imaging modalities, such as SWI will greatly spur both clinical diagnosis and advances in animal research allowing effective evaluation of potential therapies.