Computed diffusion-weighted imaging in patients with transient neurovascular symptoms with and without ischemic infarction.

PURPOSE: Detection of ischemic lesions in patients with transient neurovascular symptoms is relevant for the estimation of the risk of a subsequent stroke and etiological classification. To improve detection rates, different technical approaches have been used, such as diffusion-weighted imaging (DWI) with high b-values or higher magnetic field strength. Here, we sought to investigate the value of computed DWI (cDWI) with high b-values in these patients.

METHODS: From an MRI report database we identified patients with transient neurovascular symptoms who underwent repeated MRI including DWI. cDWI was calculated with a monoexponential model with high b-values (2000, 3000, and 4000 s/mm2 ) and compared to the routinely used standard DWI with regard to presence of ischemic lesions and lesion detectability.

RESULT: Overall 33 patients with transient neurovascular symptoms (71 [IQR 57-83.5] years; 21 [63.6%] male) were included. On DWI, acute ischemic lesions were observed in 22 (78.6%). Acute ischemic lesions were observed in 17 (51.5%) patients on initial DWI, and in 26 (78.8%) patients on follow-up DWI. Lesion detectability was rated significantly better on cDWI at 2000 s/mm2 compared to standard DWI. In 2 (9.1%) patients, cDWI at 2000 s/mm2 revealed an acute ischemic lesion proven on follow-up standard DWI which was not detected with certainty on the initial standard DWI.

CONCLUSION: cDWI might be a valuable addition to routinely acquired standard DWI in patients with transient neurovascular symptoms since its use might result in improved ischemic lesion detection. A b-value of 2000 s/mm2 seems most promising for clinical practice.