Evaluation of multiple deep neural networks for detection of intracranial dural arteriovenous fistula on susceptibility weighted angiography imaging.

BACKGROUND: The natural history of intracranial dural arteriovenous fistula (DAVF) is variable and early diagnosis is crucial in order to positively impact the clinical course of aggressive DAVF. Artificial intelligence (AI) based techniques can be promising in this regard, and in this study, we used various deep neural network (DNN) architectures to determine whether DAVF could be reliably identified on susceptibility-weighted angiography images (SWAN). MATERIALS AND METHODS: A total of 3965 SWAN image slices from 30 digital subtraction angiographically proven DAVF patients and 4380 SWAN image slices from 40 age-matched patients with normal MRI findings as control group were included. The images were categorized as either DAVF or normal and the data was trained using various DNN such as VGG-16, EfficientNet-B0, and ResNet-50. RESULTS: Various DNN architectures showed the accuracy of 95.96% (VGG-16), 91.75% (EfficientNet-B0), and 86.23% (ResNet-50) on the SWAN image dataset. ROC analysis yielded an area under the curve of 0.796 (p < .001), best for VGG-16 model. Criterion of seven consecutive positive slices for DAVF diagnosis yielded a sensitivity of 74.68% with a specificity of 69.15%, while setting eight slices improved the sensitivity to above 80.38%, with a decrease of specificity up to 56.38%. Based on seven consecutive positive slices criteria, EfficientNet-B0 yielded a sensitivity of 73.21% with a specificity of 45.92% and ResNet-50 yielded a sensitivity of 72.39% with a specificity of 67.42%. CONCLUSION: This study shows that DNN can extract discriminative features of SWAN for the classification of DAVF from normal with good accuracy, reasonably good sensitivity and specificity.

Transvenous embolization along with intraprocedural image fusion technique for complex intracranial dural arteriovenous fistula.

PURPOSE: This study aimed to investigate the efficacy and safety of an intraprocedural image fusion technique using flat-panel detector computed tomography-based rotational angiography (FDCT-RA) and image fusion (IF) for the transvenous approach in treating intracranial dural arteriovenous fistulas (dAVFs). METHODS: A retrospective review was conducted on patients who underwent transvenous embolization for dural AVFs. The patients were classified into two groups according to the treatment technique used: the FDCT-RA and IF technique group and the conventional technique group. The primary outcomes assessed were the angiographic and clinical outcomes, complications, fluoroscopy time, and radiation exposure. Univariate analyses were performed to compare the two treatment modalities. RESULTS: Eighty-six patients with intracranial dAVFs were treated with transvenous embolization (TVE), of which 37 patients underwent transvenous approach with flat-panel detector computed tomography-based rotational angiography (FDCT-RA) and image fusion (IF) technique used. The FDCT-RA and IF group showed difference in the location of dAVFs, occlusion state of the sinus, and access routes in comparison to the conventional treatment group. The FDCT-RA and IF technique was predominantly used for dAVFs involving the anterior condylar confluence and cavernous sinus with ipsilateral inferior petrosal sinus (IPS) occlusion. Patients treated with this technique demonstrated a higher rate of complete occlusion (91.9%, n = 34) compared to those treated with the conventional technique (79.6%, n = 39), but this difference was not statistically significant (p = 0.136). Although the implementation of this technique during the treatment procedure showed a tendency to decrease both fluoroscopy duration and radiation dose, the observed results did not reach statistical significance (p = 0.315, p = 0.130). CONCLUSION: The intraprocedural image fusion technique using FDCT-RA for transvenous treatment of intracranial dAVFs could provide help in treatment of dAVFs of certain locations or access routes. It might provide aid in microcatheter navigation, without increasing the radiation exposure and fluoroscopy time.

Research on segmentation method of brain vessel based on deep learning / 医疗卫生装备

Objective To propose a deep learning-based cerebrovascular segmentation method to solve the problems of magnetic resonance angiography(MRA)image auto segmentation due to some tiny or overlapped vessels.Methods Generative adversarial networks(GAN)consisting of multiple generators and discriminators were used to construct a brain vessel segmen-tation model(BVSM).Firstly,the feature fusion and attention mechanism modules were introduced into the generator network to segment and extract the patient's MRA images;secondly,the discriminator network judged the gap between the brain vessel segmentation results respectively by the generator network and the expert's manual operation,so as to optimize the generator network continuously to obtain realistic segmentation images;finally,the MIDAS dataset was used to design ablation experi-ments to compare the cerebrovascular segmentation results of BVSM with the original model(RVGAN retinal vascular gene-rative adversarial network model),the RVGAN+Attention model incorporated with the attention module and the RVGAN+slice-level feature aggregation(SFA)model with the SFA module in terms of Dice coefficient,accuracy,sensitivity and AUC.Results The BVSM behaved better than the RVGAN,RVGAN+Attention and RVGAN+SFA models with Dice coefficient being 87.2％,accuracy being 88.3％,sensitivity being 86.3％and AUC being 0.942.Conclusion The method proposed facilitates the observation of cerebrovascular structure with high accuracy,and provides an auxiliary means for diagnosing cerebrovascular diseases.[Chinese Medical Equipment Journal,2023,44(9):1-7]

Dense biased networks with deep priori anatomy and hard region adaptation: Semi-supervised learning for fine renal artery segmentation.

Fine renal artery segmentation on abdominal CT angiography (CTA) image is one of the most important tasks for kidney disease diagnosis and pre-operative planning. It will help clinicians locate each interlobar artery's blood-feeding region via providing the complete 3D renal artery tree masks. However, it is still a task of great challenges due to the large intra-scale changes, large inter-anatomy variation, thin structures, small volume ratio and small labeled dataset of the fine renal artery. In this paper, we propose the first semi-supervised 3D fine renal artery segmentation framework, DPA-DenseBiasNet, which combines deep prior anatomy (DPA), dense biased network (DenseBiasNet) and hard region adaptation loss (HRA): 1) Based on our proposed dense biased connection, the DenseBiasNet fuses multi-receptive field and multi-resolution feature maps for large intra-scale changes. This dense biased connection also obtains a dense information flow and dense gradient flow so that the training is accelerated and the accuracy is enhanced. 2) DPA features extracted from an autoencoder (AE) are embedded in DenseBiasNet to cope with the challenge of large inter-anatomy variation and thin structures. The AE is pre-trained (unsupervised) by numerous unlabeled data to achieve the representation ability of anatomy features and these features are embedded in DenseBiasNet. This process will not introduce incorrect labels as optimization targets and thus contributes to a stable semi-supervised training strategy that is suitable for sensitive thin structures. 3) The HRA selects the loss value calculation region dynamically according to the segmentation quality so the network will pay attention to the hard regions in the training process and keep the class balanced. Experiments demonstrated that DPA-DenseBiasNet had high predictive accuracy and generalization with the Dice coefficient of 0.884 which increased by 0.083 compared with 3D U-Net (Çiçek et al., 2016). This revealed our framework with great potential for the 3D fine renal artery segmentation in clinical practice.

Improving the Otsu method for MRA image vessel extraction via resampling and ensemble learning.

Accurate extraction of vessels plays an important role in assisting diagnosis, treatment, and surgical planning. The Otsu method has been used for extracting vessels in medical images. However, blood vessels in magnetic resonance angiography (MRA) image are considered as a sparse distribution. Pixels on vessels in MRA image are considered as an imbalanced data in classification of vessels and non-vessel tissues. To extract vessels accurately, a novel method using resampling technique and ensemble learning is proposed for solving the imbalanced classification problem. Each pixel is sampled multiple times through multiple local patches within the image. Then, vessel or non-vessel tissue is determined by the ensemble voting mechanism via a p-tile algorithm. Experimental results show that the proposed method is able to outperform the traditional Otsu method by extracting vessels in MRA images more accurately.

Consistency between Targets Delineated by Angiography, Computed Tomography, and Magnetic Resonance Imaging in Stereotactic Radiosurgery for Arteriovenous Malformation.

BACKGROUND: Target identification is important for radiosurgery for arteriovenous malformations (AVMs). Targets defined by different imaging modalities may be inconsistent in practice. OBJECTIVES: The goal of this study is to review and analyze the consistency between targets defined by different imaging modalities in radiosurgery for AVMs. METHODS: From March 2007 to June 2011, AVM patients for radiosurgery whose targets were delineated by angiography/computed tomography (CT)/magnetic resonance imaging (MRI) were reviewed. Spetzler-Martin grades, hemorrhage history, and treatment volumes were checked. Dice similarity coefficients (DSCs) between targets were calculated and analyzed. RESULTS: Twenty-three patients were enrolled. The mean DSCs were between 0.37 and 0.51 for targets by different modalities. There was no significant difference in DSCs regarding Spetzler-Martin grades and hemorrhage history. For CT-delineated target volumes <3 cm3, MRI-delineated target volumes <5 cm3, and angiography-delineated target volumes <2 cm3, the DSCs between the different image modalities were significantly decreased. CONCLUSIONS: Consistency between targets delineated using different image modalities was likely to be unsatisfactory and worsen significantly in niduses with volumes <5 cm3. An iterative multimodality approach to confirm the delineated targets of AVMs is suggested to be indispensable for robust treatment in radiosurgery.

Automatic coronary artery segmentation based on multi-domains remapping and quantile regression in angiographies.

Coronary artery disease has become the most dangerous diseases to human life. And coronary artery segmentation is the basis of computer aided diagnosis and analysis. Existing segmentation methods are difficult to handle the complex vascular texture due to the projective nature in conventional coronary angiography. Due to large amount of data and complex vascular shapes, any manual annotation has become increasingly unrealistic. A fully automatic segmentation method is necessary in clinic practice. In this work, we study a method based on reliable boundaries via multi-domains remapping and robust discrepancy correction via distance balance and quantile regression for automatic coronary artery segmentation of angiography images. The proposed method can not only segment overlapping vascular structures robustly, but also achieve good performance in low contrast regions. The effectiveness of our approach is demonstrated on a variety of coronary blood vessels compared with the existing methods. The overall segmentation performances si, fnvf, fvpf and tpvf were 95.135%, 3.733%, 6.113%, 96.268%, respectively.

In vivo tracking of human adipose-derived stem cells labeled with ferumoxytol in rats with middle cerebral artery occlusion by magnetic resonance imaging.

Ferumoxytol, an iron replacement product, is a new type of superparamagnetic iron oxide approved by the US Food and Drug Administration. Herein, we assessed the feasibility of tracking transplanted human adipose-derived stem cells labeled with ferumoxytol in middle cerebral artery occlusion-injured rats by 3.0 T MRI in vivo. 1 × 10(4) human adipose-derived stem cells labeled with ferumoxytol-heparin-protamine were transplanted into the brains of rats with middle cerebral artery occlusion. Neurologic impairment was scored at 1, 7, 14, and 28 days after transplantation. T2-weighted imaging and enhanced susceptibility-weighted angiography were used to observe transplanted cells. Results of imaging tests were compared with results of Prussian blue staining. The modified neurologic impairment scores were significantly lower in rats transplanted with cells at all time points except 1 day post-transplantation compared with rats without transplantation. Regions with hypointense signals on T2-weighted and enhanced susceptibility-weighted angiography images corresponded with areas stained by Prussian blue, suggesting the presence of superparamagnetic iron oxide particles within the engrafted cells. Enhanced susceptibility-weighted angiography image exhibited better sensitivity and contrast in tracing ferumoxytol-heparin-protamine-labeled human adipose-derived stem cells compared with T2-weighted imaging in routine MRI.