R

Fig. 28a-c. Fusiform ectasia (arrow) of the supraclinoid internal carotid artery with involvment of the middle and anterior cerebral artery. The 3D TOF MRA (a) reveals fusiform ectasia of the internal carotid artery and its bifurcation. The CE MRA image (b) shows reduced saturation effect and better delineation of the malformation, and allows 3D reconstruction (c)

Posterior Communicating Artery Mra

Fig. 29a-d. Acute subarchnoid hemorrhage (arrow) at CT (a). The 3D TOF MRA image (b) reveals a small aneurysm (arrow) of the internal carotid artery at the origin of the posterior communicating artery. The CE MRA image (c) better delineates the aneurysm morphology and with 3D reconstruction (d) the relationship with the posterior communicating artery is apparent either be spontaneous or occur in response to a direct trauma to the artery. Intracranially, dissecting aneurysms frequently occur near fractures or after a vessel strikes the cerebral falce [76]. Mycotic, infectious or inflammatory aneurysms occur when pathogenic agents attack the intima, for example, in cases of septic emboli which form at a vascular bifurcation, or during propagation of the infection through the vasa-vasorum. Inflammatory processes destroy the adventitia and the muscolaris mucosa. The vessel then dilates as a result of luminal pressure [77,78].

The particular hemodynamic conditions that are present within an aneurysm are responsible for its growth and rupture. Flow patterns within "berry" aneurysms are characterized by three zones: an inflow zone at the distal margin of the neck and wall of the sac; an outflow zone at the proximal margin of the neck and wall of the sac; and a central slow flow zone where the saturation of moving spins reduces the signal intensity of the aneurysm on 3D TOF MRA acquisitions [79].

3D TOF MRA is currently the non-invasive screening tool of choice for the detection of in-tracranial aneurysms. The sensitivity of this tech-

nique varies from 83% to 97% when the size of the aneurysm is greater than 5 mm, however, for smaller aneurysms the sensitivity decreases markedly [80-85]. The MRA exam must be technically accurate and must favor spatial resolution. Therefore 3D TOF MRA is more sensitive than 3D PC MRA in cases of cerebral aneurysm, especially for lesions less than 5 mm in size [34]. Furthermore, 3D TOF MRA requires markedly less time for the same acquisition volume. However, in cases of SAH, even 3D TOF MRA may require longer acquisition times because the patients are frequently irritable and unstable and hence not totally collaborative. Consequently, the best technique and parameters for each patient are often dictated by the best compromise between acquisition time and spatial resolution (Fig. 29a-d).

The introduction of TOF techniques, such as MOTSA combined with TONE has permitted slice thicknesses to be reduced to 0.75 mm, thereby minimizing saturation effects in the smaller arterial vessels. The scan times of 3D TOF MRA can be reduced by a further 40% by utilizing a section-interpolation technique [83].

3D TOF MRA should be performed within the b c

Caput Listeria

Fig. 30a-d. Axial T1-weighted (a) and T2-weighted (b) SE images reveal a partially thrombosed left intracavernous ICA aneurysm (black arrows). Spin saturation phenomena in the patent portion of the aneurysm results in an underestimation of the size of the residual aneurysm on the 3D TOF MRA image (c). MIP reconstruction underestimated the size of the residual aneurysm. The oblique DSA angiogram (d) acquired in the early arterial phase shows the real size of the partially thrombosed aneurysm (black arrow)

Fig. 30a-d. Axial T1-weighted (a) and T2-weighted (b) SE images reveal a partially thrombosed left intracavernous ICA aneurysm (black arrows). Spin saturation phenomena in the patent portion of the aneurysm results in an underestimation of the size of the residual aneurysm on the 3D TOF MRA image (c). MIP reconstruction underestimated the size of the residual aneurysm. The oblique DSA angiogram (d) acquired in the early arterial phase shows the real size of the partially thrombosed aneurysm (black arrow)

first three days of a SAH. After this time the formation of methemoglobin hampers the detection of smaller blood vessels and reduces the likelihood of seeing a small aneurysm. An alternative to MRA in patients with an acute SAH is CTA which has a sensitivity of approximately 83% for the detection of aneurysms as small as 3 mm. Unfortunately, CTA requires the patient to be exposed to at least 100 ml of iodinated contrast medium and to a considerable amount of radiation. Furthermore, the postprocessing required to obtain satisfactory images that isolate the arterial circulation is elaborate and time consuming, especially for the skull base [81].

Intra-aneurysmal flow dynamics imply that the amount of spin saturation is proportional to the size of the aneurysm, resulting in an underestimation of aneurysm size when 3D TOF MRA is used (Fig 30a-d). 3D PC MRA is affected by this phenomenon to a lesser extent. In the case of a partially thrombozed aneurysm, 3D PC MRA can demonstrate the residual lumen in which blood flow is still present. This may not be possible with 3D TOF MRA if methemaglobin is present in the thrombus as it will appear hyperintense and thus indistin guishable from the residual lumen.

For the study of large and giant aneurysms 3D CE MRA is frequently the technique of choice since it can depict the vessel lumen in a similar manner to that of CTA and DSA, even despite the lower spatial resolution of the 3D CE MRA technique. An advantage of 3D CE MRA is that MIP reconstructions are able to depict an aneurysm from numerous angles, permitting accurate visualization of the neck and of the relationship with the parent vessel (Fig. 31a-e). This is particularly useful for pre-treatment evaluation of the malformation when rotational angiography is not available. Another advantage of 3D CE MRA is that it is able to distinguish between possible thrombi and residual lumen: the signal intensity of the thrombus is always lower than that of the circulating blood in which contrast agent is present [34].

Studies comparing 3D CE MRA and 3D TOF MRA for the detection of intracranial aneurysms have indicated that the former technique is more sensitive. For example, sensitivity and specificity values of 100% and 94%, respectively, have been reported for 3D CE MRA [34]. The very short acquisition time makes this technique more feasible

Fig. 31a-e. Large aneurysm (arrow) of the right internal carotid artery at the origin of the posterior cerebral artery evident at DSA (a). 3D TOF MRA (b) shows the dilatation (arrow) but underestimates its size due to saturation. This is clearly apparent on the axial source image (arrow i n c). CE MRA (d) better delineates the malformation and permits 3D reconstruction (e), which clearly depicts the lesion and reveals its relationship to the origin of the posterior communicating artery

Fig. 31a-e. Large aneurysm (arrow) of the right internal carotid artery at the origin of the posterior cerebral artery evident at DSA (a). 3D TOF MRA (b) shows the dilatation (arrow) but underestimates its size due to saturation. This is clearly apparent on the axial source image (arrow i n c). CE MRA (d) better delineates the malformation and permits 3D reconstruction (e), which clearly depicts the lesion and reveals its relationship to the origin of the posterior communicating artery for use in emergency conditions such as in cases of SAH. Preliminary studies suggest that 3D CE MRA is comparable to DSA and superior to 3D TOF MRA for the demonstration of "berry-like" aneurysms (Table 10) [54,80-96]. Nevertheless, 3D CE MRA has not yet been accepted as a routine diagnostic modality in patients with SAH, since the achievable sensitivity does not yet compare with that of DSA. On the other hand, it is frequently used as a screening modality in at-risk patients or when the presence of aneurysm is suspected from CT or other MR examinations. Whichever technique is employed, the field of view must be positioned to cover the intracranial arterial circulation from the origin of PICA at the vertebral arteries to the pericallosal arteries, both of which are possible sites of aneurysms.

The inferior spatial resolution of 3D CE MRA compared to DSA remains, for now, the major disadvantage of this technique. However, the use of

Table 10. Accuracy of MRA versus DSA for diagnostic imaging of aneurysms of the intracranial arterial vessels

Author

MRA Technique

# Pts

Sensitivity

Specificity

Ross, 1990 [92]

cine volume gradient-echo

28

Cine only 67% Cine+partitions+spin-echo 86%

Stock, 1995 [54]

MTS VFAE MRA

50

83%

98%

Kadota, 1997 [90]

Conventional MRA

Magnetization transfer contrast and TONE MRA

< 5mm 71% 5§mm 100%

Strotzer, 1998 [95]

3D FISP 2D FLASH 3D TONE

40

94%

Metens, 2000 [34]

3D CE T1w MRA MT TONE PC

70%

94% 100% 100%

Mallouhi, 2003 [96]

VR-TOF MIP-TOF VR-CE MIP-CE

82

90.7% 83.7% 86% 86%

Okahara, 2002 [89]1

3D TOF MRA

82

Neuroradiologists 79% Experienced neurosurgeons 75% Resident radiologists 63%

1 lower for smaller aneurysms (<3 mm in maximum diameter) for multiple aneurysms and/or for aneurysms located in the ICA and ACA

larger matrix sizes and isotropic voxel acquisition may improve results.

An accepted application of MRA is in the follow up of coiled aneurysms. Several studies have demonstrated the feasibility and sensitivity of 3D TOF MRA for the detection of aneurysm recanal-ization [97, 98]. Major disadvantages of this technique however are the relatively low spatial resolution and the presence in nearly 10% of cases of susceptibility artifacts related to the coils. Recently,

MRA techniques with very short TR and TE values have been shown to reduce the number of susceptibility artifacts [99]. Moreover, the use of 3D Ce MRA with elliptic centric acquisition, seems to have better sensitivity compared to 3D TOF MRA for the demonstration of aneurysm patency (Fig. 32a-e).

Fig. 32a-e. Selective DSA (a) reveals a giant aneurysm of the left internal carotid artery. The selective DSA acquired post-treatment (b) suggests complete occlusion of the aneurysm. Almost complete occlusion is similarly suggested by the 3D TOF MRA (c). Conversely, the CE MRA image (d) reveals the presence of a little residual flow (arrow) at the neck of the aneurysm. The corresponding 3D TOF MRA (e) acquired at 3T confirms the observation made in (d) of only incomplete occlusion of the aneurysm (arrow)

Vascular Malformations

The term vascular malformation is a generic term describing various congenital lesions that can be differentiated using McCormick's classification [100]:

1. Pial (parenchymal), dural or mixed arterial-venous malformations (AVM)

2. Venous angioma

3. Cavernous angioma

4. Capillary telangiectasia

Arterial-Venous Malformations (AVM)

AVM are congenital abnormalities of vascular development and differentiation. Pial AVM have variable dimensions and can occur anywhere in the brain or spine. They are composed of complex accumulations of vascular canals which have incomplete vascular walls that permit direct communication between the arterial and venous compartments, thereby by-passing the capillary-venous districts. The principle characteristic of pial AVM is the presence of a nidus consisting of numerous afferent and efferent vascular canals (afferent and ef ferent feeders). The size of the nidus is used to classify the AVM and to determine the appropriate course of treatment: small (<3 cm), medium (between 3 and 6 cm),and large (> 6 cm) [ 101]. For example, in most cases, radiotherapy is only successful if the nidus has a diameter of less than 3 cm.

The cerebral tissue that surrounds an AVM is generally involved and can demonstrate gliosis, calcifications, and hemosiderin deposits. These alterations can explain the frequent occurance of epilepsy or neurological deficit in patients with pial AVM. The annual probability of bleeding or other neurological complications in patients with AVM varies from 2 to 4% [102].

Dural AVM represent 15% of all AVM and are diagnosed predominantly in adults. They can be distinguished as dural arterial-venous malformations (DAVM) and dural arterial-venous fistulas (DAVF).

DAVM consist of a network of ectasic arterial and venous vessels contained within a venous sinus. They are for the most part acquired, often due to the progressive dilation of arterial-venous micro-fistulas that re-canalize a thrombus of a venous sinus. The transverse and sigmoid sinuses, and to a lesser extent the cavernous sinus, are most

Table 11. Accuracy of MRA versus DSA for diagnostic imaging of vascular malformations of the intracranial arterial vessels

Author

MRA Technique

# Pts

Vessels

Sensitivity

Specificity

Stock, 1995 [54]

MTS VFAE MRA

50

100%

100%

Mallouhi, 2002 [57]

CE MRA TOF MRA

82

100% 64%

Noguchi, 2004 [108]

3D TOF

15

Intracranial dural fistulas

100%1 76%2

100% 86%

1 multiple high-intensity curvilinear or nodular structures adjacent to the sinus wall

2 high-intensity areas in the venous sinus

frequently involved. Conversely, the sagittal sinus is rarely involved. DAVM have a tendency to increase in size due to the hemodynamic alterations that they induce in the venous districts.

DAVF are often single and consist of a vessel (rarely more than one) that directly connects an artery with a vein or, more frequently, with a venous sinus, often at a high flow rate.

The symptoms associated with DAVM and DAVF depend upon the site involved and are usually linked to compression of the cranial nerves or to venous hypertension (as occurs when an AVM is located within the cavernous sinus).

In most cases both conventional MRI and MRA permit the diagnosis and characterization of AVM. In pial AVM, conventional MR can show the presence of abnormal vessels and associated parenchy-mal alterations, such as gliosis or the lack of he-mosiderin in chronic bleeding. T2-weighted MR images in particular are able to demonstrate the presence of flow-void in high flow vessels. MRA is able to complement conventional MR by clearly showing the presence of a nidus which characterizes pial or parenchymal AVM. Conventional MR is less useful in identifying DAVF however especially of the posterior cranial fossa. Carotid-cavernous fistulas can be seen directly or due to the presence of indirect signs such as dilation of the superior ophthalmic vein and enlargement of the cavernous sinus (Fig. 33a-e).

The MRA technique most often used is 3D TOF, however, 2D/3D PC MRA with and without contrast agent is also used to characterize flow in abnormal vessels [103-107]. The accuracy values of MRA reported in the literature for the detection of AVM are shown in Table 11 [94,96,108] (Fig. 34a-e).

Unfortunately, information concerning the he-modynamics of AVM is often lacking with MRA. For example, despite the panoramic visualization achievable on MRA, accurate definition of the afferent and efferent vessels of the nidus is often poor. This is partly due to the low spatial resolution compared to DSA. The lack of temporal reso lution is another important limitation of MRA since hemodynamic data are essential for correct pre-treatment evaluation of AVM and other developmental abnormalities (DVA). To overcome these limitations, various authors have investigated the use of CE MRA [41, 109] and CE MR DSA techniques [86] for the characterization of AVM. The introduction of ultra fast sequences has permitted a dynamic or time resolved approach which permits differentiation of the efferent from the afferent vessels.

Among the advantages of the newer 2D CE MR DSA techniques is a temporal resolution of 1-2 sec which has proven satisfactory for studies of the intracranial arterial-venous circulation [86]. Although 3D CE MRA has also been used to study the vascular architecture of AVM, the longer acquisition times, although still less than a minute, are generally insufficient to permit satisfactory differentiation of the afferent vessels from the efferent vessels [41].

Although DSA is still the most accurate technique for the identification and characterization of AVM, it is impractical and dangerous to use routinely to monitor the progression of the pathology. In these cases 3D TOF MRA [110] and, more recently, CE MRA have proven to be valid alternatives for the planning and follow-up of radio-surgery.

Venous Angioma

Venous malformations (venous angioma) consist of deep small estasic venous vessels with a radial pattern which feed an elastic transcortical vein or, more rarely, a subependimal vein (Fig. 35a-c). They are not vascular malformations but anatomical variations or DVA [111]. DVA are most frequently seen in the frontal or cerebellar semioval centers. Since bleeding is comparatively rare, the cerebral tissue surrounding these malformations is normal. A slight probability of bleeding is present for the

Image Normal Mra Cavernous CarotidMixed Cavernous Dva Anomaly Brain

Fig. 33a-e. Right carotid cavernous fistula. The 3D TOF MRA (a) shows the presence of abnormal flow (arrow) in the right cavernous sinus and in the superior ophthalmic vein. The corresponding 3D PC MRA (b) reveals enlargment of the cavernous sinus and of the right ophthalmic vein (arrow) while the 2D phase PC MRA image (c) demonstrates the anomalous flow direction of the superior ophthalmic vein. The time-resolved CE MRA dynamic series (d) acquired in the axial plane with a temporal resolution of 1.2 seconds, reveals early enhancement of the enlarged cavernous sinus and of the ophthalmic vein. Selective DSA (e) of the right internal carotid artery confirms the presence of the direct fistula (arrow)

Carotid DsaDural Arterio Malformation

Fig. 34a-h. A 32-year-old woman with arterio-venous malformation (AVM). The frontal DSA angiograms (a, b) acquired in the arterial and early venous phases reveal a pial (parenchymal) AVM with multiple feeding vessels and large cortical nidus (a) and large cortical draining veins (blackarrowsin b). The 2D PC MRA image (c) shows the cortical draining veins (white arrows), while the nidus and arterial feeders are better shown on the 3D TOF MRA image (d). Lateral DSA angiograms (e, f) acquired in the arterial and early venous phases reval the nidus, and the arterial and venous (blackarrowin f) feeders. The 2D PC MRA lateral view image (g) shows one of the large venous feeders (blackarrow). However, this is not seen on 3D TOF MRA (h)

Fig. 34a-h. A 32-year-old woman with arterio-venous malformation (AVM). The frontal DSA angiograms (a, b) acquired in the arterial and early venous phases reveal a pial (parenchymal) AVM with multiple feeding vessels and large cortical nidus (a) and large cortical draining veins (blackarrowsin b). The 2D PC MRA image (c) shows the cortical draining veins (white arrows), while the nidus and arterial feeders are better shown on the 3D TOF MRA image (d). Lateral DSA angiograms (e, f) acquired in the arterial and early venous phases reval the nidus, and the arterial and venous (blackarrowin f) feeders. The 2D PC MRA lateral view image (g) shows one of the large venous feeders (blackarrow). However, this is not seen on 3D TOF MRA (h)

Venous Angioma

Fig. 35a-b. Right frontal venous angioma. The anomaly is clearly apparent on the T2-weighted (a) and contrast-enhanced T1-weighted (b) images as a linear vascular structure characterized by flow void and enhancement (arrows). The 3D TOF MRA image (c) acquired after injection of contrast agent reveals enhancement of the abnormal drainage due to slow flow

Fig. 35a-b. Right frontal venous angioma. The anomaly is clearly apparent on the T2-weighted (a) and contrast-enhanced T1-weighted (b) images as a linear vascular structure characterized by flow void and enhancement (arrows). The 3D TOF MRA image (c) acquired after injection of contrast agent reveals enhancement of the abnormal drainage due to slow flow mixed form which is associated with venous and cavernous angiomas [112].

Most DVA are asymptomatic and are typically incidental findings. DVA are rarely visible on 3D TOF MRA because the flow within the vessels is very slow with characteristics similar to those of venous vessels. 3D PC MRA acquired with VENC values that range from 10-20 cm/sec are capable of revealing DVA, although the best techniques for identifying DVA are 2D and 3D CE MRA. Both are capable of visualizing the typical characteristics of DVA which include small deep veins flowing to the Medusa head (Caput medusa).

Cavernous Angioma

Cavernous angiomas are vascular malformations which can be seen not only in the brain and spine but also in the vertebra, liver and spleen. Cavernous angiomas are predominantly localized in the cerebral hemispheres but can also be seen in the cerebral trunk and the cerebellum. The presence of hemosiderin at the periphery of an an-gioma reflects the possibility that red blood cells are deposited outside the incomplete blood vessel for reasons other than bleeding. The probability of bleeding for cavernous angioma, seen on MRI as a breaking of the hemosiderin shell, is very low [113]. Cavernous angiomas are mostly asymptomatic but may induce neurological symptoms due to compression of cerebral structures. Epilepsy is the most common symptom when the lesion involves the cerebral cortex. Multiple cavernous an-giomas with a positive family history are consid ered hereditary with an autosomal dominant characteristic [114].

The typical appearance of cavernous angiomas on MR include a hyperintense nucleus on T1- and T2-weighted images which reflects the presence of methemoglobin, and a hypointense delimiting ring on T2-weighted images (above all on T2w GRE images) due to the presence of hemosiderin. The most sensitive MR imaging sequence for multiple angiomas is T2w GRE, which should always be included in the scan protocol.

Cavernous angiomas were historically referred to as cryptic angiomas because they were not visible with cerebral angiography. Usually cerebral an-giography is not requested except in cases in which a mixed form is suspected. For the same reason, MRA is not typically employed in cases of cavernous angioma. In the case of 3D TOF MRA, the acquisition may be affected by the short T1 of the methaemoglobin component of the malformation.

Capillary Telangectasia

Capillary telangectasia is defined as a mass of dilated capillaries which have an incomplete vascular wall often surrounded by normal cerebral tissue and frequently located in the pons [115]. Glio-sis or hemosiderin may be present in the area surrounding the lesion. Patients with capillary telangectasia are usually asymptomatic and diagnosis is typically incidental on contrast enhanced MR examinations. Most capillary telangectasia are too small to be detected by MRA and in most cases by cerebral angiography too.

Cerebral Angiography Ipss
a

Fig. 36a-b. 3D TOF MRA of the intracranial arteries at 1.5T (a) and 3T (b). An overall higher signal of all intracranial arteries with better visualization of cortical branches is evident at 3T

Fig. 37a-b. 3D TOF MRA at 1.5T (a) shows the presence of a very small aneurysm of the cavernous ICA (arrow). However, better depiction of the aneurysm (arrow) is achieved on 3D TOF MRA at 3T (b) due to improvements in signal and spatial resolution
Essentials of Human Physiology

Essentials of Human Physiology

This ebook provides an introductory explanation of the workings of the human body, with an effort to draw connections between the body systems and explain their interdependencies. A framework for the book is homeostasis and how the body maintains balance within each system. This is intended as a first introduction to physiology for a college-level course.

Get My Free Ebook


Post a comment