Comparison of magnetic resonance angiography scans on 1.5, 3, and 7 Tesla units: a quantitative study of 3-dimensional cerebrovasculature

eICAB: A novel deep learning pipeline for Circle of Willis multiclass segmentation and analysis

BACKGROUND

The accurate segmentation, labeling and quantification of cerebral blood vessels on MR imaging is important for basic and clinical research, yet results are not generalizable, and often require user intervention. New methods are needed to automate this process.

PURPOSE

To automatically segment, label and quantify Circle of Willis (CW) arteries on Magnetic Resonance Angiography images using deep convolutional neural networks.

MATERIALS AND METHODS

MRA images were pooled from three public and private databases. A total of 116 subjects (mean age 56 years ± 21 [standard deviation]; 72 women) were used to make up the training set (N=101) and the testing set (N=15). In each image, fourteen arterial segments making up or surrounding the CW were manually annotated and validated by a clinical expert. Convolutional neural network (CNN) models were trained on a training set to be finally combined in an ensemble to develop eICAB. Model performances were evaluated using (1) quantitative analysis (dice score on test set) and (2) qualitative analysis (external datasets, N=121). The reliability was assessed using multiple MRAs of healthy participants (ICC of vessel diameters and volumes on test-retest).

RESULTS

Qualitative analysis showed that eICAB correctly predicted the large, medium and small arteries in 99±0.4%, 97±1% and 88±7% of all images, respectively. For quantitative assessment, the average dice score coefficients for the large (ICAs, BA), medium (ACAs, MCAs, PCAs-P2), and small (AComm, PComm, PCAs-P1) vessels were 0.76±0.07, 0.76±0.08 and 0.41±0.27, respectively. These results were similar and, in some cases, statistically better (p<0.05) than inter-expert annotation variability and robust to image SNR. Finally, test-retest analysis showed that the model yielded high diameter and volume reliability (ICC=0.99).

CONCLUSION

We have developed a quick and reliable open-source CNN-based method capable of accurately segmenting and labeling the CW in MRA images. This method is largely independent of image quality. In the future, we foresee this approach as a critical step towards fully automated analysis of MRA databases in basic and clinical research.

7T versus 3T MR Angiography to Assess Unruptured Intracranial Aneurysms

BACKGROUND AND PURPOSE

Aneurysm size and neck measurements are important for treatment decisions. The introduction of 7T magnetic resonance angiography (MRA) led to new possibilities assessing aneurysm morphology and flow due to the higher signal-to-noise ratio. However, it is unknown if the size measurements on 7T MRA are similar to those on the standard 3T MRA. This study aimed to compare aneurysm size measurements between 7T and 3T MRA.

METHODS

We included 18 patients with 22 aneurysms who underwent both 3T and 7T MRA. Three acquisition protocols were compared: 3T time of flight (TOF), 7T TOF, and 7T contrast-enhanced MRA. Each aneurysm on each protocol was measured by at least two experienced neuroradiologists. Subsequently, the differences were evaluated using scatterplots and the intraclass correlation coefficients (ICC) of agreement.

RESULTS

There was a good agreement among the neuroradiologists for the height and width measurements (mean ICC: .78-.93); the neck measurements showed a moderate agreement with a mean ICC of .57-.72. Between the MR acquisition protocols, there was a high agreement for all measurements with a mean ICC of .81-.96. Measurement differences between acquisition protocols (0-2.9 mm) were in the range of the differences between the neuroradiologists (0-3.6 mm).

CONCLUSION

Our study showed that 7T MRA, both nonenhanced and contrast-enhanced, has a high agreement in aneurysm size measurements compared to 3T. This suggests that 7T is useful for reliable aneurysm size assessment.

Plaque Distribution Correlates With Morphology of Lenticulostriate Arteries in Single Subcortical Infarctions

Supplemental Digital Content is available in the text.

Background and Purpose:

We aimed to use novel whole-brain vessel-wall magnetic resonance imaging (WB-VWI) to investigate the association between plaque distribution of middle cerebral artery (MCA) and morphological changes of the lenticulostriate arteries (LSAs) in single subcortical infarctions.

Methods:

Forty single subcortical infarction patients with no relevant MCA disease on magnetic resonance angiography were prospectively enrolled. Plaque location in the MCA was dichotomized as proximal (located adjacent to the LSA origin) or distal (located distal to the LSA origin) on whole-brain vessel-wall magnetic resonance imaging. The MCAs with proximal plaques were divided into the symptomatic and asymptomatic side, and asymptomatic side MCAs without proximal plaques were the control group. The morphological characteristics of the LSAs and features of proximal plaques were analyzed.

Results:

A total of 71 MCAs in 40 patients were analyzed (31 on the symptomatic side, 22 on the asymptomatic side, and 18 in the control group). Superior-wall plaques of MCAs were observed more frequently on the symptomatic side than the asymptomatic side (45.2% versus 9.1%, P=0.005). The wall area index, plaque burden, and remodeling index did not differ significantly between the symptomatic and asymptomatic side. The number of LSA branches was smaller (P=0.011) in the symptomatic side (5.48±1.88) compared with the control group (6.83±1.92). The symptomatic side exhibited shorter average length of the LSAs (23.23±3.44 versus 25.75±3.76 mm, P=0.025) and shorter average distance of the LSAs (16.47±3.11 versus 21.53±4.76 mm, P<0.001) compared with the asymptomatic side.

Conclusions:

Superiorly distributed MCA plaques at the LSA origin are closely associated with morphological changes of the LSA in symptomatic MCAs, suggesting that the distribution, rather than the inherent features of plaques, determines the occurrence of single subcortical infarctions. Our findings provide insight into the etiologic mechanism of branch atheromatous disease in single subcortical infarctions.

Neuroimaging of Intracerebral Hemorrhage

Intracerebral hemorrhage (ICH) accounts for 10% to 20% of strokes worldwide and is associated with high morbidity and mortality rates. Neuroimaging is indispensable for rapid diagnosis of ICH and identification of the underlying etiology, thus facilitating triage and appropriate treatment of patients. The most common neuroimaging modalities include noncontrast computed tomography (CT), CT angiography (CTA), digital subtraction angiography, and magnetic resonance imaging (MRI). The strengths and disadvantages of each modality will be reviewed. Novel technologies such as dual-energy CT/CTA, rapid MRI techniques, near-infrared spectroscopy, and automated ICH detection hold promise for faster pre- and in-hospital ICH diagnosis that may impact patient management.

High resolution time-of-flight MR-angiography at 7 T exploiting VERSE saturation, compressed sensing and segmentation

BACKGROUND

3D Time-of-Flight (TOF) MR-angiography (MRA) substantially benefits from ultra-high magnetic field strengths (≥7 T) due to increased Signal-to-Noise ratio and improved contrast. However, high-resolution TOF-MRA usually requires long acquisition times. In addition, specific absorption rate constraints limit the choice of optimal pulse sequence parameters, especially if venous saturation is employed.

PURPOSE

To implement and evaluate an arterial TOF-MRA for accelerated high-resolution angiography at ultra-high magnetic field strength.

FIELD STRENGTHS/SEQUENCE

7 T modified gradient-echo TOF sequence including venous saturation using Variable-Rate Selective Excitation (VERSE), Compressed Sensing (CS) and sparse application of saturation pulses, called segmentation, were included for acceleration.

ASSESSMENT

To analyze the acceleration techniques all volunteers were examined with the same protocols. CS with different sampling patterns and regularization factors as well as segmentation were applied for acceleration. For comparison, conventional acceleration techniques were applied (GRAPPA PAT 3 and Partial Fourier (6/8 in slice/phase encoding)). Images were co-registered and 40 mm thick transversal maximum intensity projections were created to calculate the relative number of vessels. To analyze the visibility of small vessels, the lenticulostriate arteries (LSA) were examined. This was done via multiscale vessel enhancement filtering in a VOI and quantification via Fiji ImageJ as well as qualitatively evaluation by two radiologists. Additionally, the venous/arterial vessel-to-background ratios (vVBR/aVBR) were calculated for chosen protocols.

RESULTS

For the acceleration of a high resolution TOF-MRA (0.31 mm isotropic), under-sampling of 9.6 showed aliasing artifacts, whereas 7.2 showed no aliasing. The regularization factor R had a strong impact on the image quality according to smoothing (R = 0.01 to R = 0.005) and noise (R = 0.0005 to R = 0.00005). With the alternating sampling patterns it was shown that the k-space center should not be under-sampled too much. Additionally segmentation could be verified to be feasible for stronger acceleration with sufficient venous suppression.

CONCLUSION

The combination of several independent techniques (VERSE, CS with acceleration factor 7.2, R = 0.001, Poisson disc radius of 80%, 3 segments) enables the application of high-resolution (0.31 mm isotropic) TOF-MRA with venous saturation at 7 T in clinical time settings (TA ≈ 5 min) and within the SAR limits.

7-T MRI in Cerebrovascular Diseases: Challenges to Overcome and Initial Results

Magnetic resonance imaging (MRI) plays a key role in the investigation of cerebrovascular diseases. Compared with computed tomography (CT) and digital subtraction angiography (DSA), its advantages in diagnosing cerebrovascular pathology include its superior tissue contrast, its ability to visualize blood vessels without the use of a contrast agent, and its use of magnetic fields and radiofrequency pulses instead of ionizing radiation. In recent years, ultrahigh field MRI at 7 tesla (7 T) has shown promise in the diagnosis of many cerebrovascular diseases. The increased signal-to-noise ratio (SNR; 2.3x and 4.7x increase compared with 3 and 1.5 T, respectively) and contrast-to-noise ratio (CNR) at this higher field strength can be exploited to obtain a higher spatial resolution and higher lesion conspicuousness, enabling assessment of smaller brain structures and lesions. Cerebrovascular diseases can be assessed at different tissue levels; for instance, changes of the arteries feeding the brain can be visualized to determine the cause of ischemic stroke, regional changes in brain perfusion can be mapped to predict outcome after revascularization, and tissue damage, including old and recent ischemic infarcts, can be evaluated as a marker of ischemic burden. For the purpose of this review, we will discriminate 3 levels of assessment of cerebrovascular diseases using MRI: Pipes, Perfusion, and Parenchyma (3 Ps). The term Pipes refers to the brain-feeding arteries from the heart and aortic arch, upwards to the carotid arteries, vertebral arteries, circle of Willis, and smaller intracranial arterial branches. Perfusion is the amount of blood arriving at the brain tissue level, and includes the vascular reserve and perfusion territories. Parenchyma refers to the acute and chronic burden of brain tissue damage, which includes larger infarcts, smaller microinfarcts, and small vessel disease manifestations such as white matter lesions, lacunar infarcts, and microbleeds. In this review, we will describe the key developments in the last decade of 7-T MRI of cerebrovascular diseases, subdivided for these 3 levels of assessment.

Recent applications of UHF-MRI in the study of human brain function and structure: a review

The increased availability of ultra-high-field (UHF) MRI has led to its application in a wide range of neuroimaging studies, which are showing promise in transforming fundamental approaches to human neuroscience. This review presents recent work on structural and functional brain imaging, at 7 T and higher field strengths. After a short outline of the effects of high field strength on MR images, the rapidly expanding literature on UHF applications of blood-oxygenation-level-dependent-based functional MRI is reviewed. Structural imaging is then discussed, divided into sections on imaging weighted by relaxation time, including quantitative relaxation time mapping, phase imaging and quantitative susceptibility mapping, angiography, diffusion-weighted imaging, and finally magnetization-transfer imaging. The final section discusses studies using the high spatial resolution available at UHF to identify explicit links between structure and function. Copyright © 2015 John Wiley & Sons, Ltd.

Investigation of the saturation pulse artifact in non-enhanced MR angiography of the lower extremity arteries at 7 Tesla

When performing non-enhanced time-of-flight MR angiography of the lower extremity arteries at 7 T with cardiac triggering, the acquisition time is a crucial consideration. Therefore, in previous studies, saturation RF pulses were applied only every second TR. In the axial source images a slight artifact with an appearance similar to aliasing could be observed. The purpose of this study was to investigate the origin of this artifact. The reason for the artifact is supposed to be related to the two effective TRs during acquisition caused by the sparsely applied saturation RF pulse. Several sequence variants were simulated and implemented within the sequence source code to examine this hypothesis. An adaptation of the excitation flip angles for each TR as well as a correction factor for the k-space data was calculated. Additionally, a different ordering of the k-space data during acquisition was implemented as well as the combination of the latter with the k-space correction factor. The observations from the simulations were verified using both a static and a flow phantom and, finally, in a healthy volunteer using the same measurement setup as in previous volunteer and patient studies. Of all implemented techniques, only the reordering of the k-space was capable of suppressing the artifact almost completely at the cost of creating a ringing artifact. The phantom measurements showed the same results as the simulations and could thus confirm the hypothesis regarding the origin of the artifact. This was additionally verified in the healthy volunteer. The origin of the artifact could be confirmed to be the periodic signal variation caused by two effective TRs during acquisition.

High-resolution postcontrast time-of-flight MR angiography of intracranial perforators at 7.0 Tesla

Background and Purpose

Different studies already demonstrated the benefits of 7T for precontrast TOF-MRA in the visualization of intracranial small vessels. The aim of this study was to assess the performance of high-resolution 7T TOF-MRA after the administration of a gadolinium-based contrast agent in visualizing intracranial perforating arteries.

Materials and Methods

Ten consecutive patients (7 male; mean age, 50.4 ± 9.9 years) who received TOF-MRA at 7T after contrast administration were retrospectively included in this study. Intracranial perforating arteries, branching from the parent arteries of the circle of Willis, were identified on all TOF-MRA images. Provided a TOF-MRA before contrast administration was present, a direct comparison between pre- and postcontrast TOF-MRA was made.

Results

It was possible to visualize intracranial perforating arteries branching off from the entire circle of Willis, and their proximal branches. The posterior cerebral artery (P1 and proximal segment of P2) appeared to have the largest number of visible perforating branches (mean of 5.1 in each patient, with a range of 2–7). The basilar artery and middle cerebral artery (M1 and proximal segment M2) followed with a mean number of 5.0 and 3.5 visible perforating branches (range of 1–9 and 1–8, respectively). Venous contamination in the postcontrast scans sometimes made it difficult to discern the arterial or venous nature of a vessel.

Conclusion

High-resolution postcontrast TOF-MRA at 7T was able to visualize multiple intracranial perforators branching off from various parts of the circle of Willis and proximal intracranial arteries. Although confirmation in a larger study is needed, the administration of a contrast agent for high-resolution TOF-MRA at 7T seems to enable a better visualization of the distal segment of certain intracranial perforators.

Geometric accuracy of 3D coordinates of the Leksell stereotactic skull frame in 1.5 Tesla- and 3.0 Tesla-magnetic resonance imaging: a comparison of three different fixation screw materials

We assessed the geometric distortion of 1.5-Tesla (T) and 3.0-T magnetic resonance (MR) images with the Leksell skull frame system using three types of cranial quick fixation screws (QFSs) of different materials-aluminum, aluminum with tungsten tip, and titanium-for skull frame fixation. Two kinds of acrylic phantoms were placed on a Leksell skull frame using the three types of screws, and were scanned with computed tomography (CT), 1.5-T MR imaging and 3.0-T MR imaging. The 3D coordinates for both strengths of MR imaging were compared with those for CT. The deviations of the measured coordinates at selected points (x = 50, 100 and 150; y = 50, 100 and 150) were indicated on different axial planes (z = 50, 75, 100, 125 and 150). The errors of coordinates with QFSs of aluminum, tungsten-tipped aluminum, and titanium were <1.0, 1.0 and 2.0 mm in the entire treatable area, respectively, with 1.5 T. In the 3.0-T field, the errors with aluminum QFSs were <1.0 mm only around the center, while the errors with tungsten-tipped aluminum and titanium were >2.0 mm in most positions. The geometric accuracy of the Leksell skull frame system with 1.5-T MR imaging was high and valid for clinical use. However, the geometric errors with 3.0-T MR imaging were larger than those of 1.5-T MR imaging and were acceptable only with aluminum QFSs, and then only around the central region.

Improved cerebral time-of-flight magnetic resonance angiography at 7 Tesla--feasibility study and preliminary results using optimized venous saturation pulses

Purpose

Conventional saturation pulses cannot be used for 7 Tesla ultra-high-resolution time-of-flight magnetic resonance angiography (TOF MRA) due to specific absorption rate (SAR) limitations. We overcome these limitations by utilizing low flip angle, variable rate selective excitation (VERSE) algorithm saturation pulses.

Material and Methods

Twenty-five neurosurgical patients (male n = 8, female n = 17; average age 49.64 years; range 26–70 years) with different intracranial vascular pathologies were enrolled in this trial. All patients were examined with a 7 Tesla (Magnetom 7 T, Siemens) whole body scanner system utilizing a dedicated 32-channel head coil. For venous saturation pulses a 35° flip angle was applied. Two neuroradiologists evaluated the delineation of arterial vessels in the Circle of Willis, delineation of vascular pathologies, presence of artifacts, vessel-tissue contrast and overall image quality of TOF MRA scans in consensus on a five-point scale. Normalized signal intensities in the confluence of venous sinuses, M1 segment of left middle cerebral artery and adjacent gray matter were measured and vessel-tissue contrasts were calculated.

Results

Ratings for the majority of patients ranged between good and excellent for most of the evaluated features. Venous saturation was sufficient for all cases with minor artifacts in arteriovenous malformations and arteriovenous fistulas. Quantitative signal intensity measurements showed high vessel-tissue contrast for confluence of venous sinuses, M1 segment of left middle cerebral artery and adjacent gray matter.

Conclusion

The use of novel low flip angle VERSE algorithm pulses for saturation of venous vessels can overcome SAR limitations in 7 Tesla ultra-high-resolution TOF MRA. Our protocol is suitable for clinical application with excellent image quality for delineation of various intracranial vascular pathologies.

A super-resolution ultrasound method for brain vascular mapping

PURPOSE

High-resolution vascular imaging has not been achieved in the brain due to limitations of current clinical imaging modalities. The authors present a method for transcranial ultrasound imaging of single micrometer-size bubbles within a tube phantom.

METHODS

Emissions from single bubbles within a tube phantom were mapped through an ex vivo human skull using a sparse hemispherical receiver array and a passive beamforming algorithm. Noninvasive phase and amplitude correction techniques were applied to compensate for the aberrating effects of the skull bone. The positions of the individual bubbles were estimated beyond the diffraction limit of ultrasound to produce a super-resolution image of the tube phantom, which was compared with microcomputed tomography (micro-CT).

RESULTS

The resulting super-resolution ultrasound image is comparable to results obtained via the micro-CT for small tissue specimen imaging.

CONCLUSIONS

This method provides superior resolution to deep-tissue contrast ultrasound and has the potential to be extended to provide complete vascular network imaging in the brain.

Sequence comparison for non-enhanced MRA of the lower extremity arteries at 7 Tesla

In this study three sequences for non-contrast-enhanced MRA of the lower extremity arteries at 7T were compared. Cardiac triggering was used with the aim to reduce signal variations in the arteries. Two fast single-shot 2D sequences, a modified Ultrafast Spoiled Gradient Echo (UGRE) sequence and a variant of the Quiescent-Interval Single-Shot (QISS) sequence were triggered via phonocardiogram and compared in volunteer examinations to a non-triggered 2D gradient echo (GRE) sequence. For image acquisition, a 16-channel transmit/receive coil and a manually positionable AngioSURF table were used. To tackle B₁ inhomogeneities at 7T, Time-Interleaved Acquisition of Modes (TIAMO) was integrated in GRE and UGRE. To compare the three sequences quantitatively, a vessel-to-background ratio (VBR) was measured in all volunteers and stations. In conclusion, cardiac triggering was able to suppress flow artifacts satisfactorily. The modified UGRE showed only moderate image artifacts. Averaged over all volunteers and stations, GRE reached a VBR of 4.18±0.05, UGRE 5.20±0.06, and QISS 2.72±0.03. Using cardiac triggering and TIAMO imaging technique was essential to perform non-enhanced MRA of the lower extremities vessels at 7T. The modified UGRE performed best, as observed artifacts were only moderate and the highest average VBR was reached.

Non-enhanced MR imaging of cerebral aneurysms: 7 Tesla versus 1.5 Tesla

PURPOSE

To prospectively evaluate 7 Tesla time-of-flight (TOF) magnetic resonance angiography (MRA) in comparison to 1.5 Tesla TOF MRA and 7 Tesla non-contrast enhanced magnetization-prepared rapid acquisition gradient-echo (MPRAGE) for delineation of unruptured intracranial aneurysms (UIA).

MATERIAL AND METHODS

Sixteen neurosurgical patients (male n = 5, female n = 11) with single or multiple UIA were enrolled in this trial. All patients were accordingly examined at 7 Tesla and 1.5 Tesla MRI utilizing dedicated head coils. The following sequences were obtained: 7 Tesla TOF MRA, 1.5 Tesla TOF MRA and 7 Tesla non-contrast enhanced MPRAGE. Image analysis was performed by two radiologists with regard to delineation of aneurysm features (dome, neck, parent vessel), presence of artifacts, vessel-tissue-contrast and overall image quality. Interobserver accordance and intermethod comparisons were calculated by kappa coefficient and Lin's concordance correlation coefficient.

RESULTS

A total of 20 intracranial aneurysms were detected in 16 patients, with two patients showing multiple aneurysms (n = 2, n = 4). Out of 20 intracranial aneurysms, 14 aneurysms were located in the anterior circulation and 6 aneurysms in the posterior circulation. 7 Tesla MPRAGE imaging was superior over 1.5 and 7 Tesla TOF MRA in the assessment of all considered aneurysm and image quality features (e.g. image quality: mean MPRAGE7T: 5.0; mean TOF7T: 4.3; mean TOF1.5T: 4.3). Ratings for 7 Tesla TOF MRA were equal or higher over 1.5 Tesla TOF MRA for all assessed features except for artifact delineation (mean TOF7T: 4.3; mean TOF1.5T 4.4). Interobserver accordance was good to excellent for most ratings.

CONCLUSION

7 Tesla MPRAGE imaging demonstrated its superiority in the detection and assessment of UIA as well as overall imaging features, offering excellent interobserver accordance and highest scores for all ratings. Hence, it may bear the potential to serve as a high-quality diagnostic tool for pretherapeutic assessment and follow-up of untreated UIA.

Digital reconstruction and morphometric analysis of human brain arterial vasculature from magnetic resonance angiography

Characterization of the complex branching architecture of cerebral arteries across a representative sample of the human population is important for diagnosing, analyzing, and predicting pathological states. Brain arterial vasculature can be visualized by magnetic resonance angiography (MRA). However, most MRA studies are limited to qualitative assessments, partial morphometric analyses, individual (or small numbers of) subjects, proprietary datasets, or combinations of the above limitations. Neuroinformatics tools, developed for neuronal arbor analysis, were used to quantify vascular morphology from 3T time-of-flight MRA high-resolution (620 μm isotropic) images collected in 61 healthy volunteers (36/25 F/M, average age=31.2 ± 10.7, range=19-64 years). We present in-depth morphometric analyses of the global and local anatomical features of these arbors. The overall structure and size of the vasculature did not significantly differ across genders, ages, or hemispheres. The total length of the three major arterial trees stemming from the circle of Willis (from smallest to largest: the posterior, anterior, and middle cerebral arteries; or PCAs, ACAs, and MCAs, respectively) followed an approximate 1:2:4 proportion. Arterial size co-varied across individuals: subjects with one artery longer than average tended to have all other arteries also longer than average. There was no net right-left difference across the population in any of the individual arteries, but ACAs were more lateralized than MCAs. MCAs, ACAs, and PCAs had similar branch-level properties such as bifurcation angles. Throughout the arterial vasculature, there were considerable differences between branch types: bifurcating branches were significantly shorter and straighter than terminating branches. Furthermore, the length and meandering of bifurcating branches increased with age and with path distance from the circle of Willis. All reconstructions are freely distributed through a public database to enable additional analyses and modeling (cng.gmu.edu/brava).

Pediatric high-field magnetic resonance imaging

High-field 3 T magnetic resonance (MR) imaging provides greater signal-to-noise ratio (SNR) compared with 1.5 T systems. Various MR imaging clinical applications in children can benefit from improvements resulting from this increased SNR. High-resolution imaging of the brain, arterial spin labeling perfusion imaging, diffusion imaging, MR spectroscopy, and imaging of small anatomic parts are some areas in which these improvements can increase our clinical diagnostic capabilities. However, challenges inherent to 3 T imaging become more relevant in children. The use of 3 T imaging in children has allowed better diagnostic efficacy in neuroimaging, but certain technique modifications may be required for optimal imaging.

The Role of 3 Tesla MRA in the Detection of Intracranial Aneurysms

Intracranial aneurysms constitute a common pathological entity, affecting approximately 1–8% of the general population. Their early detection is essential for their prompt treatment. Digital subtraction angiography is considered the imaging method of choice. However, other noninvasive methodologies such as CTA and MRA have been employed in the investigation of patients with suspected aneurysms. MRA is a noninvasive angiographic modality requiring no radiation exposure. However, its sensitivity and diagnostic accuracy were initially inadequate. Several MRA techniques have been developed for overcoming all these drawbacks and for improving its sensitivity. 3D TOF MRA and contrast-enhanced MRA are the most commonly employed techniques. The introduction of 3 T magnetic field further increased MRA's sensitivity, allowing detection of aneurysms smaller than 3 mm. The development of newer MRA techniques may provide valuable information regarding the flow characteristics of an aneurysm. Meticulous knowledge of MRA's limitations and pitfalls is of paramount importance for avoiding any erroneous interpretation of its findings.

The predictive value of 3D time-of-flight MR angiography in assessment of brain arteriovenous malformation obliteration after radiosurgery

BACKGROUND AND PURPOSE

The purpose of radiosurgery of bAVMs is complete angiographic obliteration of its nidus. We assessed the diagnostic accuracy of 1.5T T2-weighted MR imaging and TOF-MRA images for detecting nidus obliteration after radiosurgery.

MATERIALS AND METHODS

The pre- and postradiosurgery MR images and DSA images from 120 patients who were radiosurgically treated for a bAVM were re-evaluated by 2 observers for patency of the nidus (preradiosurgery) and obliteration (postradiosurgery: final follow-up MR imaging), by using a 3-point scale of confidence. Consensus reading of the DSA after radiosurgery was considered the criterion standard for obliteration. Sensitivity, specificity, PPVs, and NPVs, and overall diagnostic performance by using ROC were determined.

RESULTS

Mean bAVM volume during radiosurgery was 3.4 mL (95% CI, 2.6-4.3 mL). Sixty-six patients (55%) had undergone previous endovascular embolization. The mean intervals between radiosurgery and follow-up MR imaging and for DSA, respectively, were 35.6 months (95% CI, 32.3-38.9 months) and 42.1 months (95% CI, 40.3-44.0 months). With ROC, an area under curve of 0.81-0.83 was found. PPVs of final follow-up MR-imaging for definitive obliteration varied between 0.89 [corrected] and 0.95. NPV was 0.52 [corrected] . An average false-positive rate, meaning overestimation of nidus obliteration of 0.10 [corrected] and an average false-negative rate, meaning underestimation of nidus obliteration of 0.42 [corrected] were found.

CONCLUSIONS

MRA is insufficient to diagnose obliteration in the follow-up of bAVMs after radiosurgery. A remaining nidus diameter <10 mm seems to be the major limiting factor for reliable assessment of obliteration. We highly recommend follow-up DSA for definitive diagnosis of complete obliteration.