A comparison of contrast-free MRA at 3.0 T in cases of intracranial aneurysms with or without subarachnoid hemorrhage

Geometric Deep Learning Using Vascular Surface Meshes for Modality-Independent Unruptured Intracranial Aneurysm Detection

Early detection of unruptured intracranial aneurysms (UIAs) enables better rupture risk and preventative treatment assessment. UIAs are usually diagnosed on Time-of-Flight Magnetic Resonance Angiographs (TOF-MRA) or contrast-enhanced Computed Tomography Angiographs (CTA). Various automatic voxel-based deep learning UIA detection methods have been developed, but these are limited to a single modality. We propose a modality-independent UIA detection method using a geometric deep learning model with high resolution surface meshes of brain vessels. A mesh convolutional neural network with ResU-Net style architecture was used. UIA detection performance was investigated with different input and pooling mesh resolutions, and including additional edge input features (shape index and curvedness). Both a higher resolution mesh (15,000 edges) and additional curvature edge features improved performance (average sensitivity: 65.6%, false positive count/image (FPC/image): 1.61). UIAs were detected in an independent TOF-MRA test set and a CTA test set with average sensitivity of 52.0% and 48.3% and average FPC/image of 1.04 and 1.05 respectively. We provide modality-independent UIA detection using a deep-learning vascular surface mesh model with comparable performance to state-of-the-art UIA detection methods.

Super-resolution application of generative adversarial network on brain time-of-flight MR angiography: image quality and diagnostic utility evaluation

OBJECTIVES

To develop a generative adversarial network (GAN) model to improve image resolution of brain time-of-flight MR angiography (TOF-MRA) and to evaluate the image quality and diagnostic utility of the reconstructed images.

METHODS

We included 180 patients who underwent 1-min low-resolution (LR) and 4-min high-resolution (routine) brain TOF-MRA scans. We used 50 patients' datasets for training, 12 for quantitative image quality evaluation, and the rest for diagnostic validation. We modified a pix2pix GAN to suit TOF-MRA datasets and fine-tuned GAN-related parameters, including loss functions. Maximum intensity projection images were generated and compared using multi-scale structural similarity (MS-SSIM) and information theoretic-based statistic similarity measure (ISSM) index. Two radiologists scored vessels' visibilities using a 5-point Likert scale. Finally, we evaluated sensitivities and specificities of GAN-MRA in depicting aneurysms, stenoses, and occlusions.

RESULTS

The optimal model was achieved with a lambda of 1e5 and L1 + MS-SSIM loss. Image quality metrics for GAN-MRA were higher than those for LR-MRA (MS-SSIM, 0.87 vs. 0.73; ISSM, 0.60 vs. 0.35; p.adjusted < 0.001). Vessels' visibility of GAN-MRA was superior to LR-MRA (rater A, 4.18 vs. 2.53; rater B, 4.61 vs. 2.65; p.adjusted < 0.001). In depicting vascular abnormalities, GAN-MRA showed comparable sensitivities and specificities, with greater sensitivity for aneurysm detection by one rater (93% vs. 84%, p < 0.05).

CONCLUSIONS

An optimized GAN could significantly improve the image quality and vessel visibility of low-resolution brain TOF-MRA with equivalent sensitivity and specificity in detecting aneurysms, stenoses, and occlusions.

KEY POINTS

• GAN could significantly improve the image quality and vessel visualization of low-resolution brain MR angiography (MRA). • With optimally adjusted training parameters, the GAN model did not degrade diagnostic performance by generating substantial false positives or false negatives. • GAN could be a promising approach for obtaining higher resolution TOF-MRA from images scanned in a fraction of time.

Prevalence of unruptured intracranial aneurysms: impact of different definitions - the Tromsø Study

BACKGROUND

Management of incidental unruptured intracranial aneurysms (UIAs) remains challenging and depends on their risk of rupture, estimated from the assumed prevalence of aneurysms and the incidence of aneurysmal subarachnoid haemorrhage. Reported prevalence varies, and consistent criteria for definition of UIAs are lacking. We aimed to study the prevalence of UIAs in a general population according to different definitions of aneurysm.

METHODS

Cross-sectional population-based study using 3-dimensional time-of-flight 3 Tesla MR angiography to identify size, type and location of UIAs in 1862 adults aged 40-84 years. Size was measured as the maximal distance between any two points in the aneurysm sac. Prevalence was estimated for different diameter cutoffs (≥1, 2 and 3 mm) with and without inclusion of extradural aneurysms.

RESULTS

The overall prevalence of intradural saccular aneurysms ≥2 mm was 6.6% (95% CI 5.4% to 7.6%), 7.5% (95% CI 5.9% to 9.2%) in women and 5.5% (95% CI 4.1% to 7.2%) in men. Depending on the definition of an aneurysm, the overall prevalence ranged from 3.8% (95% CI 3.0% to 4.8%) for intradural aneurysms ≥3 mm to 8.3% (95% CI 7.1% to 9.7%) when both intradural and extradural aneurysms ≥1 mm were included.

CONCLUSION

Prevalence in this study was higher than previously observed in other Western populations and was substantially influenced by definitions according to size and extradural or intradural location. The high prevalence of UIAs sized <5 mm may suggest lower rupture risk than previously estimated. Consensus on more robust and consistent radiological definitions of UIAs is warranted.

Cost-Effective Analysis of Different Diagnostic Strategies in Screening for Aneurysms After Spontaneous Subarachnoid Hemorrhage

PURPOSE

With an increasing number of patients being treated by coiling for aneurysms, using computed tomographic angiography (CTA) or magnetic resonance angiography (MRA) as an initial screening test becomes less important because digital subtraction angiography (DSA) is the first step of endovascular treatment procedure. The objective of this study is to investigate whether CTA or MRA remained to be the optimal screening strategy in patients with spontaneous subarachnoid hemorrhage (SAH).

METHODS

CTA, MRA, and DSA strategy were evaluated in a decision tree model created with TreeAge Pro Suite 2011. Input parameters were derived from published literature and our institutional database. Base case and sensitivity analyses were conducted to assess the cost-effectiveness of each strategy. A Monte Carlo simulation was performed with all parameters ranged among their distributions to evaluate the validation of results.

RESULTS

The base case scenario showed that MRA was the most cost-effective strategy. Using a willingness-to-pay threshold of ¥70,892/quality adjusted life year, MRA remained to be most cost-effective when its sensitivity is >0.907. DSA was not cost-effective compared to CTA or MRA unless over 91.56% of patients were treated by coiling. The Monte Carlo simulation reported DSA not to be a cost-effective strategy at willingness-to-pay of ¥70,892 in 99.99% of the iterations.

CONCLUSION

DSA is not cost-effective compared to CTA or MRA and should not be used as the initial diagnostic tool for spontaneous SAH.

MRI and MR angiography evaluation of pulsatile tinnitus: A focused, physiology-based protocol

BACKGROUND AND PURPOSE

Pulsatile tinnitus (PT) is the subjective sensation of a pulse-synchronous sound, most often due to a cerebrovascular etiology. PT can severely impact quality of life and may indicate a life-threatening process, yet a timely and accurate diagnosis can often lead to effective treatment. Clinical assessment with a history and physical examination can often suggest a diagnosis for PT, but is rarely definitive. Therefore, PT should be evaluated with a comprehensive and targeted radiographic imaging protocol. MR imaging provides a safe and effective means to evaluate PT. Specific MR sequences may be used to highlight different elements of cerebrovascular anatomy and physiology. However, routine MR evaluation of PT must comply with economic and practical constraints, while effectively capturing both common and rarer, life-threatening etiologies of PT.

METHODS

In this state-of-the-art review, we describe our institutional MR protocol for evaluating PT.

RESULTS

This protocol includes the following dedicated sequences: time-of-flight magnetic resonance angiography; arterial spin labeling; spoiled gradient recalled acquisition in the steady state; time-resolved imaging of contrast kinetics; diffusion weighted imaging, and 3-dimensional fluid-attenuated inversion recovery.

CONCLUSIONS

We describe the physiologic and clinical rationale for including each MR sequence in a comprehensive PT imaging protocol, and detail the role of MR within the broader evaluation of PT, from clinical presentation to treatment.

The sensitivity and specificity of TOF-MRA compared with DSA in the follow-up of treated intracranial aneurysms

BACKGROUND

Time-of-flight magnetic resonance angiography (TOF-MRA) is widely used in detecting intracranial aneurysms (IA), but it is limited and controversial for use during follow-up to assess the outcome of interventional coiling or clipping surgery.

METHODS

To evaluate the specificity and sensitivity of using TOF-MRA as an imaging follow-up for IA with different treatments. A total of 280 patients with 326 treated IA underwent simultaneous TOF-MRA and digital subtraction angiography (DSA) as follow-up imaging on the same day. All images were independently reviewed by two neurosurgeons and two radiologists. The consensus evaluation of intra-arterial DSA as a reference test was used to evaluate the result of aneurysm occlusions. The aneurysmal embolization status was assessed with two ratings involving complete or incomplete occlusions. We calculated the sensitivity, specificity, negative predictive value, and positive predictive value of three-dimensional-TOF-MRA to investigate the diagnostic performance.

RESULTS

Overall sensitivity and specificity of TOF-MRA for diagnosing the remnant were 83.3% and 95.2%, respectively. The sensitivity and specificity of interventional therapy was 90.0% and 94.2%, respectively, while the clipping group showed sensitivity and specificity of 50.0% and 100%, respectively. For additional groups, involving coil only, stent-assisted, and flow diverter, the analysis of interventional therapy showed sensitivities and specificities of 100.0% and 90.1%, 66.7% and 95.1%, and 91.7% and 100%, respectively.

CONCLUSIONS

TOF-MRA can be used as a first-line noninvasive imaging modality during follow-up, especially for the patients treated with a pipeline embolization device and coils only. But it may not be enough for clipped aneurysms.

Deep Learning Approach for Generating MRA Images From 3D Quantitative Synthetic MRI Without Additional Scans

OBJECTIVES

Quantitative synthetic magnetic resonance imaging (MRI) enables synthesis of various contrast-weighted images as well as simultaneous quantification of T1 and T2 relaxation times and proton density. However, to date, it has been challenging to generate magnetic resonance angiography (MRA) images with synthetic MRI. The purpose of this study was to develop a deep learning algorithm to generate MRA images based on 3D synthetic MRI raw data.

MATERIALS AND METHODS

Eleven healthy volunteers and 4 patients with intracranial aneurysms were included in this study. All participants underwent a time-of-flight (TOF) MRA sequence and a 3D-QALAS synthetic MRI sequence. The 3D-QALAS sequence acquires 5 raw images, which were used as the input for a deep learning network. The input was converted to its corresponding MRA images by a combination of a single-convolution and a U-net model with a 5-fold cross-validation, which were then compared with a simple linear combination model. Image quality was evaluated by calculating the peak signal-to-noise ratio (PSNR), structural similarity index measurements (SSIMs), and high frequency error norm (HFEN). These calculations were performed for deep learning MRA (DL-MRA) and linear combination MRA (linear-MR), relative to TOF-MRA, and compared with each other using a nonparametric Wilcoxon signed-rank test. Overall image quality and branch visualization, each scored on a 5-point Likert scale, were blindly and independently rated by 2 board-certified radiologists.

RESULTS

Deep learning MRA was successfully obtained in all subjects. The mean PSNR, SSIM, and HFEN of the DL-MRA were significantly higher, higher, and lower, respectively, than those of the linear-MRA (PSNR, 35.3 ± 0.5 vs 34.0 ± 0.5, P < 0.001; SSIM, 0.93 ± 0.02 vs 0.82 ± 0.02, P < 0.001; HFEN, 0.61 ± 0.08 vs 0.86 ± 0.05, P < 0.001). The overall image quality of the DL-MRA was comparable to that of TOF-MRA (4.2 ± 0.7 vs 4.4 ± 0.7, P = 0.99), and both types of images were superior to that of linear-MRA (1.5 ± 0.6, for both P < 0.001). No significant differences were identified between DL-MRA and TOF-MRA in the branch visibility of intracranial arteries, except for ophthalmic artery (1.2 ± 0.5 vs 2.3 ± 1.2, P < 0.001).

CONCLUSIONS

Magnetic resonance angiography generated by deep learning from 3D synthetic MRI data visualized major intracranial arteries as effectively as TOF-MRA, with inherently aligned quantitative maps and multiple contrast-weighted images. Our proposed algorithm may be useful as a screening tool for intracranial aneurysms without requiring additional scanning time.