CAVE: Cerebral artery-vein segmentation in digital subtraction angiography.

Cerebral X-ray digital subtraction angiography (DSA) is a widely used imaging technique in patients with neurovascular disease, allowing for vessel and flow visualization with high spatio-temporal resolution. Automatic artery-vein segmentation in DSA plays a fundamental role in vascular analysis with quantitative biomarker extraction, facilitating a wide range of clinical applications. The widely adopted U-Net applied on static DSA frames often struggles with disentangling vessels from subtraction artifacts. Further, it falls short in effectively separating arteries and veins as it disregards the temporal perspectives inherent in DSA. To address these limitations, we propose to simultaneously leverage spatial vasculature and temporal cerebral flow characteristics to segment arteries and veins in DSA. The proposed network, coined CAVE, encodes a 2D+time DSA series using spatial modules, aggregates all the features using temporal modules, and decodes it into 2D segmentation maps. On a large multi-center clinical dataset, CAVE achieves a vessel segmentation Dice of 0.84 (±0.04) and an artery-vein segmentation Dice of 0.79 (±0.06). CAVE surpasses traditional Frangi-based k-means clustering (P < 0.001) and U-Net (P < 0.001) by a significant margin, demonstrating the advantages of harvesting spatio-temporal features. This study represents the first investigation into automatic artery-vein segmentation in DSA using deep learning. The code is publicly available at https://github.com/RuishengSu/CAVE_DSA.

Towards quantitative digital subtraction perfusion angiography: An animal study.

BACKGROUND: X-ray digital subtraction angiography (DSA) is the imaging modality for peri-procedural guidance and treatment evaluation in (neuro-) vascular interventions. Perfusion image construction from DSA, as a means of quantitatively depicting cerebral hemodynamics, has been shown feasible. However, the quantitative property of perfusion DSA has not been well studied. PURPOSE: To comparatively study the independence of deconvolution-based perfusion DSA with respect to varying injection protocols, as well as its sensitivity to alterations in brain conditions. METHODS: We developed a deconvolution-based algorithm to compute perfusion parametric images from DSA, including cerebral blood volume (CBV D S A $_{DSA}$ ), cerebral blood flow (CBF D S A $_{DSA}$ ), time to maximum (Tmax), and mean transit time (MTT D S A $_{DSA}$ ) and applied it to DSA sequences obtained from two swine models. We also extracted the time intensity curve (TIC)-derived parameters, that is, area under the curve (AUC), peak concentration of the curve, and the time to peak (TTP) from these sequences. Deconvolution-based parameters were quantitatively compared to TIC-derived parameters in terms of consistency upon variations in injection profile and time resolution of DSA, as well as sensitivity to alterations of cerebral condition. RESULTS: Comparing to TIC-derived parameters, the standard deviation (SD) of deconvolution-based parameters (normalized with respect to the mean) are two to five times smaller, indicating that they are more consistent across different injection protocols and time resolutions. Upon ischemic stroke induced in a swine model, the sensitivities of deconvolution-based parameters are equal to, if not higher than, those of TIC-derived parameters. CONCLUSIONS: In comparison to TIC-derived parameters, deconvolution-based perfusion imaging in DSA shows significantly higher quantitative reliability against variations in injection protocols across different time resolutions, and is sensitive to alterations in cerebral hemodynamics. The quantitative nature of perfusion angiography may allow for objective treatment assessment in neurovascular interventions.

Cost-efficient anthropomorphic head phantom for quantitative image quality assessment in cone beam CT.

In this study, a novel anthropomorphic head phantom for quantitative image quality assessment in cone beam computed tomography (CBCT) is proposed. The phantom is composed of tissue equivalent materials (TEMs) which are suitable for cost-efficient fabrication methods such as silicone casting and 3D printing. A monocalcium phosphate/gypsum mixture (MCPHG), nylon and a silyl modified polymer gel (SMP) are proposed as bone, muscle and brain equivalent materials respectively. The TEMs were evaluated for their radiodensity in terms of Hounsfield Units (HU) and their x-ray scatter characteristics. The median radiodensity and inter quartile range (IQR) of the MCPHG and SMP were found to be within the range of the theoretical radiodensity for bone and brain tissue: 922 (IQR = 156) and 47 (IQR = 7) HU respectively. The median radiodensity of nylon was slightly outside of the HU range of muscle tissue, but within the HU range of a combination of muscle and adipose tissue: -18 (IQR = 40) HU. The median ratios between the measured scatter characteristics and simulated tissues were between 0.84 and 1.13 (IQR between 0.05 and 0.14). The preliminary results of this study show that the proposed design and TEMs are potentially suitable for the fabrication of a cost-efficient anthropomorphic head phantom for quantitative image quality assessment in CT or CBCT.

Spatio-temporal deep learning for automatic detection of intracranial vessel perforation in digital subtraction angiography during endovascular thrombectomy.

Intracranial vessel perforation is a peri-procedural complication during endovascular therapy (EVT). Prompt recognition is important as its occurrence is strongly associated with unfavorable treatment outcomes. However, perforations can be hard to detect because they are rare, can be subtle, and the interventionalist is working under time pressure and focused on treatment of vessel occlusions. Automatic detection holds potential to improve rapid identification of intracranial vessel perforation. In this work, we present the first study on automated perforation detection and localization on X-ray digital subtraction angiography (DSA) image series. We adapt several state-of-the-art single-frame detectors and further propose temporal modules to learn the progressive dynamics of contrast extravasation. Application-tailored loss function and post-processing techniques are designed. We train and validate various automated methods using two national multi-center datasets (i.e., MR CLEAN Registry and MR CLEAN-NoIV Trial), and one international multi-trial dataset (i.e., the HERMES collaboration). With ten-fold cross-validation, the proposed methods achieve an area under the curve (AUC) of the receiver operating characteristic of 0.93 in terms of series level perforation classification. Perforation localization precision and recall reach 0.83 and 0.70 respectively. Furthermore, we demonstrate that the proposed automatic solutions perform at similar level as an expert radiologist.

Effects of changing physiologic conditions on the in vivo quantification of hemodynamic variables in cerebral aneurysms treated with flow diverting devices.

Quantifying the hemodynamic environment within aneurysms and its change after deployment of flow diverting devices is important to assess the device efficacy and understand their long-term effects. The purpose of this study was to estimate deviations in the quantification of the relative change of hemodynamic variables during flow diversion treatment of cerebral aneurysms due to changing physiologic flow conditions. Computational fluid dynamics calculations were carried out on three patient-specific geometries. Three flow diverters were virtually implanted in each geometry and simulations were performed under five pulsatile flow conditions. Hemodynamic variables including aneurysm inflow rate, mean velocity, shear rate, and wall shear stress were quantified before and after stenting. Deviations in the relative change of these variables due to varying flow conditions were calculated. The results indicate that a change in the mean flow of the parent artery of approximately 30-50% can induce large deviations in the relative change of hemodynamic variables in the range of 30-80%. Thus, quantification of hemodynamic changes during flow diversion must be carried out carefully. Variations in the inflow conditions during the procedure may induce large deviations in the quantification of these changes.

Fused magnetic resonance angiography and 2D fluoroscopic visualization for endovascular intracranial neuronavigation.

Advanced transluminal neurovascular navigation is an indispensable image-guided method that allows for real-time navigation of endovascular material in critical neurovascular settings. Thus far, it has been primarily based on 2D and 3D angiography, burdening the patient with a relatively high level of iodinated contrast. However, in the patients with renal insufficiency, this method is no longer tolerable due to the contrast load. The authors present a novel image guidance technique based on periprocedural fluoroscopic images fused with a preinterventionally acquired MRI data set. The technique is illustrated in a case in which the fused image combination was used for endovascular treatment of a giant cerebral aneurysm.

On-line multi-slice computed tomography interactive overlay with conventional X-ray: a new and advanced imaging fusion concept.

BACKGROUND: Computed tomography (CT) has revolutionized noninvasive cardiovascular evaluations. Complicated percutaneous procedures require precise imaging guidance that conventional X-ray is often unable to provide. By combining X-ray imaging with real-time, interactive, CT-based landmarks, interventional procedures could be facilitated. We describe two cases using the first CT/Live X-ray overlay in which this technology shows its potential. CASE REPORTS: A 31-year-old male with an anatomically complicated atrial septal defect (ASD) was referred for percutaneous closure. Transesophageal echocardiography (TEE) revealed an inferior location of the ASD complicated by it's proximity to a prominent Eustachian ridge. The CT was used to create a patient-specific physical model in preparation for the procedure and an in-lab real-time CT overlay allowing successful closure. A second case of a 41-year-old male with coronary artery disease status-post coronary artery bypass, aortic valve replacement (AVR), and aortic root replacement with an abnormal coronary computed tomography angiogram (CTA). In a prior procedure years ago the saphenous vein graft (SVG) to the left anterior descending artery (LAD) could not be cannulated during invasive angiography, given the patient's complicated and unusual anatomy. Using CT overlay, the superiorly and anteriorly located SVG was cannulated successfully. DISCUSSION: CT/Live X-ray overlay provided an adequate anatomical intra-procedural ASD evaluation, defect sizing, and guidance in one case and localization of an anatomically challenging graft ostium in the other case. Adding the CT landmarks as an overlay to traditional X-ray techniques provides a revolutionary and advanced imaging fusion concept that should improve procedural success.