Cerebral venous thrombosis (venous sinus thrombosis) is an elusive diagnosis because of its nonspecific presentation and its numerous predisposing causes (see the images below). It is more common than previously thought. Imaging plays a key role in the diagnosis.
Cerebral venous thrombosis often presents with hemorrhagic infarction in areas atypical for arterial vascular distribution. Magnetic resonance venography (MRV) in conjunction with conventional MRI can accurately diagnose cerebral venous thrombosis. With careful interpretation and a high degree of clinical suspicion, CT also may lead to the diagnosis. [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11]
MRI with MRV is preferred for diagnosis of cerebral venous thrombosis (venous sinus thrombosis). Clinical manifestations and physical findings may be nonspecific. The diagnosis may be made or suggested by CT brain scan before and after intravenous contrast medium injection. [12, 13, 14]
Two-dimensional time-of-flight (2D TOF) MRV is performed in the coronal plane; however, in-plane signal loss that mimics thrombosis may occur with this technique. Thus, a review of source data and conventional MRI brain scan is necessary. Phase-contrast MRV techniques may help, since small cortical venous infarcts may not be observed on 2D TOF MRV. 
The diagnosis of cerebral venous thrombosis (venous sinus thrombosis) occasionally can be made by CT scan with careful attention to the findings, which may be subtle (see the images below).
Diagnostic findings include the following:
On noncontrast CT scan, the classic finding is the delta sign, which is observed as a dense triangle (from hyperdense thrombus) within the superior sagittal sinus. However, this is not specific, since high attenuation in the healthy nonthrombosed sinus can be observed occasionally and is common in neonates because of an elevated hematocrit. 
On contrast-enhanced CT scan, the reverse delta sign (ie, empty triangle sign) can be observed in the superior sagittal sinus from enhancement of the dural leaves surrounding the comparatively less dense thrombosed sinus.
The presence of both the delta and reverse delta signs increases the likelihood of the diagnosis.
On CT brain scan, infarctions in a nonarterial distribution in the white matter and/or cortical white matter junction, often associated with hemorrhage, should suggest the possible diagnosis of venous thrombosis. Bilateral cerebral involvement can occur, including the superior cerebral white matter of the convexities from superior sagittal sinus thrombosis, or the basal ganglia and thalami from internal cerebral vein thrombosis in which the internal cerebral veins appear hyperdense in the noncontrast scan.
Indirect CT signs include focal cerebral cortical ischemia with gyral enhancement, small ventricles compressed by cerebral edema, and intense tentorial enhancement. Occasionally, the transcerebral medullary cortical veins can be observed.
The characteristic CT scan appearances and signs strongly suggest cerebral venous thrombosis, but CT scans are seldom conclusively diagnostic. Because of the subtlety of the findings, the prospective diagnosis of venous thrombosis may not be made unless a high index of suspicion is maintained during interpretation of the CT study. CT venography also can confirm the diagnosis  and does not suffer from the in-plane flow artifact that causes signal loss on 2-dimensional time-of-flight (2D TOF) magnetic resonance venography (MRV).
A false-positive delta sign may occur in a trauma setting because of an adjacent subdural hematoma. The dural sinus normally may appear hyperdense relative to adjacent tissues. Both the delta and reverse delta signs may need to be present to establish the diagnosis of cerebral venous thrombosis (venous sinus thrombosis). Occasionally, the superior sagittal sinus may bifurcate proximal to the torcular herophili, resulting in a confluens sinuum and an empty delta sign without the presence of a venous thrombosis.
In most patients, MRI brain scan with MRV is recommended to establish the diagnosis. Cerebral arteriography and venography may be necessary when MRI scan and MRV are not available.
Magnetic Resonance Imaging
MRI in conjunction with magnetic resonance venography (MRV) is both sensitive and specific enough to provide the best noninvasive method of diagnosing cerebral venous thrombosis (see the images below).
The diagnosis usually can be made without intravenous contrast, although contrast enhancement can aid in confirming the diagnosis. [5, 18, 19, 20, 21] A thrombus can be directly visualized within a vessel. Secondary venous infarctions and foci of hemorrhage can be seen with gradient-echo images. Susceptibility-induced signal loss from deoxyhemoglobin provides a basis for detection of even small foci of hemorrhage, which tend to occur in the subcortical white matter, thalami, and basal ganglia.
Parenchymal regions of T2-hyperintense signal abnormality in the distribution of the draining sinus is often observed and may be reversible, even when large. This may occur independent of recanalization of the thrombosed vessels. Dilated venous collaterals, such as transcortical medullary veins, provide indirect evidence of venous thrombosis. The appearance of intravenous thrombus on conventional MRI depends on the age of the blood clot within the vessel.
In acute venous thrombosis, loss of flow void on T1-weighted images occurs along with hypointensity on T2-weighted images, making the determination of sinus occlusion difficult. In the subacute phase, blood clot can result in loss of normal flow void on T1-weighted images and T1 hyperintensity; conversely, on T2-weighted images, blood clot can be of low signal intensity, thus mimicking flowing blood. In this instance, blood is in the intracellular methemoglobin stage.
Flow-related enhancement phenomena created by slow flow can occur in veins and cause T1 hyperintensity. To circumvent this problem, flow-sensitive imaging techniques can be used (ie, 2-dimensional time-of-flight [2D TOF] or phase-contrast MRV) to accurately assess the venous sinuses; 2D TOF MRV pulse sequence is sensitive to slow flow. Maximum signal is produced when blood flows orthogonal to the imaging plane, and since many cerebral veins course in an anteroposterior direction, coronal acquisition is often used with an inferior saturation pulse to eliminate arterial signal.
Restricted diffusion may or may not be seen in cerebral venous thrombosis and, when present, may occasionally be reversible.  Partially recanalized chronic venous sinus thrombosis has been described to demonstrate intense enhancement of the thrombosed segments following intravenous gadolinium-based contrast agents. A case report of restricted diffusion in the optic nerves bilaterally has been described in the setting of cavernous sinus thrombosis.
The role of the fibrin-specific MR contrast agent EP-2104R, which is based on gadolinium, has been postulated to bind only to fibrin and not to circulating fibrinogen and has been demonstrated in animal models to be highly selective for cerebral sinus vein thrombosis. This agent may indicate a developing role for molecular imaging. 
Pffefferkorn et al studied 32 patients with cerebral venous thrombosis, and they found that the most frequent parenchymal MRI finding was thalamic edema (69% of patients; bilateral in 47%). 
Variants of venous anatomy are common, and a hypoplastic sinus or prominent arachnoid granulations may simulate venous sinus thrombosis. With 2D TOF MRV techniques, thrombus in the intracellular or extracellular methemoglobin stage can present with increased signal and falsely simulate blood flow. Phase-contrast MRV may avoid this error.
Hypoplasia or severe attenuation of a transverse sinus, which are normal anatomic variants, may simulate venous sinus thrombosis. In-plane flow-induced signal loss in 2D TOF MRV also can mimic intravenous thrombus. Prominent arachnoid granulations may simulate thrombus. A careful review of the MRV images and conventional MRI may lead to the correct diagnosis.
In general, ultrasonography is not useful for cerebral venous thrombosis (venous sinus thrombosis). Better noninvasive modalities include MRI with MRV and CT venography. Similar to other imaging modalities, differentiation of the normal anatomic variant of hypoplastic sinus from a thrombosed sinus can be difficult. In neonates, however, the diagnosis of venous sinus thrombosis may be made by color Doppler ultrasound. An isolated case diagnosed with power Doppler was reported in the literature. [24, 25]
A published series from Germany reported 14 patients with transverse sinus thrombosis studied by transcranial color duplex ultrasonography and concluded that without contrast media administration, ultrasonography is "almost useless." 
The use of Levovist imaging agent (ie, transpulmonary stable microbubbles formed in galactose suspension) results in a signal increase of 25 dB and offers improved diagnostic capability.
Nuclear medicine has no significant role in the evaluation of cerebral venous thrombosis; however, localized reduced perfusion in the affected venous distribution can be observed. Radionuclide study may result in a false-positive finding in partial thrombosis. MR and CT venography are preferred. Because of the low spatial resolution of the radionuclide study, collateral vessels adjacent to a thrombosed venous sinus may mimic a patent sinus. Congenital variations in venous anatomy (eg, unilateral transverse sinus) are not discernible from venous sinus occlusion.
Prior to the advent of MRI, the diagnosis of cerebral venous thrombosis (venous sinus thrombosis)was confirmed by contrast arteriography.
Classic findings include the following:
Filling defects from thrombus within the venous sinus
Occlusion of a draining sinus
Secondary indirect angiographic findings are as follows:
Decreased focal venous circulation around a thrombosed venous sinus
Visualization of collateral circulation
Narrowing of arteries in the involved region
Prolonged contrast blush in the brain parenchyma
Tortuous vessels in the capillary and venous phases
Collateral flow in dilated anastomotic vessels
For severe cases of cerebral venous sinus thrombosis, intravascular thrombolytic therapy can be performed within the affected sinus if conventional heparin therapy fails.  Localized thrombolysis during the postpartum period has also been successful.  A multicenter randomized clinical trial is in progress to determine whether this therapy improves functional outcome, with more detailed information available at TO-ACT Trial.