The following information on patient history was obtained: TIA and minor strokes we classified into the following categories: retinal TIA, cerebral

TIA or stroke. Documented were the nature of the events such as visual, pure motor, pure sensory, dysarthria, dysphasia, ataxia, apraxia or combination of events. ABCD2 scores were obtained in all patients [6]. MRI findings were classified into cortical infarcts, subcortical infarcts and leucoaraiosis. Infarcts were further subdivided into recent or this website non-recent and left or right sided. The side, severity of the stenosis and presence of plaque ulceration on duplex and CTA were documented as well. Furthermore, blood pressure was documented as well as the current use of anti-thrombotic drugs or anti-coagulants. Documentation of the TCD embolus detection included: the side of insonation, the peak systolic-, mean and end-diastolic velocity, the duration of the measurement and the presence or absence of cerebral embolism by human experts. If experts found cerebral embolism the following parameters 5-FU ic50 of that embolus were noted: velocity, phase of cardiac cycle (systolic/diastolic) in which the events occurred, intensity, duration and a parameter related to the musical characteristics of the embolus (the zero-crossing index) [7]. Data of stent

procedures and surgery were prospectively documented including the occurrence of neurological or non-neurological complications. The follow-up at three month included a neurological visit at the outpatient clinic. Documented were the TIA and stroke recurrence rate. If complications had occurred in the post-operative phase of angioplasty or surgery they were evaluated including the occurrence of new medical events in the last three months. All data were stored in a downloadable

Internet based electronic management system which allowed online statistical analysis of all included case records. This data management system has been developed by Mediwebdesign© The Netherlands (http://www.mediwebdesign.nl/spi/stroke/loginreal.php). A TCD Delica 9 series (Delicate/Shenzen/China) equipped with a 2 MHz TCD transducer and a notebook PC (Acer®, Aspire 1800 Series) were used for this study. A special Delicate headband was used to hold the 2 MHz transducer, which allowed Verteporfin hands-off monitoring. The insonated artery was the middle cerebral artery at its origin, just lateral of the terminal internal carotid artery, on the ipsilateral side of the symptomatic carotid artery territory. Patients were monitored for 30 min. In case of positive embolism the other contra-lateral middle cerebral artery was examined to estimate whether the cerebral embolism was a uni-lateral or bilateral phenomenon. Insonation depth varied between 45 mm and 55 mm. Patients were asked to not speak or move their head during the monitoring session because angular or lateral probe movements may induce false positive embolic events.