USING CONVOLUTIONAL NEURAL NETWORK-BASED SEGMENTATION FOR IMAGE-BASED COMPUTATIONAL FLUID DYNAMICS SIMULATIONS OF BRAIN ANEURYSMS: INITIAL EXPERIENCE IN AUTOMATED MODEL CREATION

Automatic segmentation of intraluminal thrombosis of abdominal aortic aneurysms from CT angiography using a mixed-scale-driven multiview perception network (M2Net) model

Intraluminal thrombosis (ILT) plays a critical role in the progression of abdominal aortic aneurysms (AAA). Understanding the role of ILT can improve the evaluation and management of AAAs. However, compared with highly developed automatic vessel lumen segmentation methods, ILT segmentation is challenging. Angiographic contrast agents can enhance the vessel lumen but cannot improve boundary delineation of the ILT regions; the lack of intrinsic contrast in the ILT structure significantly limits the accurate segmentation of ILT. Additionally, ILT is not evenly distributed within AAAs; its sparsity and scattered distributions in the imaging data pose challenges to the learning process of neural networks. Thus, we propose a multiview fusion approach, allowing us to obtain high-quality ILT delineation from computed tomography angiography (CTA) data. Our multiview fusion network is named Mixed-scale-driven Multiview Perception Network (M2Net), and it consists of two major steps. Following image preprocessing, the 2D mixed-scale ZoomNet segments ILT from each orthogonal view (i.e., Axial, Sagittal, and Coronal views) to enhance the prior information. Then, the proposed context-aware volume integration network (CVIN) effectively fuses the multiview results. Using contrast-enhanced computed tomography angiography (CTA) data from human subjects with AAAs, we evaluated the proposed M2Net. A quantitative analysis shows that the proposed deep-learning M2Net model achieved superior performance (e.g., DICE scores of 0.88 with a sensitivity of 0.92, respectively) compared with other state-of-the-art deep-learning models. In closing, the proposed M2Net model can provide high-quality delineation of ILT in an automated fashion and has the potential to be translated into the clinical workflow.

Exploring a frequency-domain attention-guided cascade U-Net: Towards spatially tunable segmentation of vasculature

Developing fully automatic and highly accurate medical image segmentation methods is critically important for vascular disease diagnosis and treatment planning. Although advances in convolutional neural networks (CNNs) have spawned an array of automatic segmentation models converging to saturated high performance, none have explored whether CNNs can achieve (spatially) tunable segmentation. As a result, we propose multiple attention modules from a frequency-domain perspective to construct a unified CNN architecture for segmenting vasculature with desired (spatial) scales. The proposed CNN architecture is named frequency-domain attention-guided cascaded U-Net (FACU-Net). Specifically, FACU-Net contains two innovative components: (1) a frequency-domain-based channel attention module that adaptively tunes channel-wise feature responses and (2) a frequency-domain-based spatial attention module that enables the deep network to concentrate on foreground regions of interest (ROIs) effectively. Furthermore, we devised a novel frequency-domain-based content attention module to enhance or weaken the high (spatial) frequency information, allowing us to strengthen or eliminate vessels of interest. Extensive experiments using clinical data from patients with intracranial aneurysms (IA) and abdominal aortic aneurysms (AAA) demonstrated that the proposed FACU-Net met its design goal. In addition, we further investigated the association between varying (spatial) frequency components and the desirable vessel size/scale attributes. In summary, our preliminary findings are encouraging, and further developments may lead to deployable image segmentation models that are spatially tunable for clinical applications.

S-Net: a multiple cross aggregation convolutional architecture for automatic segmentation of small/thin structures for cardiovascular applications

With the success of U-Net or its variants in automatic medical image segmentation, building a fully convolutional network (FCN) based on an encoder-decoder structure has become an effective end-to-end learning approach. However, the intrinsic property of FCNs is that as the encoder deepens, higher-level features are learned, and the receptive field size of the network increases, which results in unsatisfactory performance for detecting low-level small/thin structures such as atrial walls and small arteries. To address this issue, we propose to keep the different encoding layer features at their original sizes to constrain the receptive field from increasing as the network goes deeper. Accordingly, we develop a novel S-shaped multiple cross-aggregation segmentation architecture named S-Net, which has two branches in the encoding stage, i.e., a resampling branch to capture low-level fine-grained details and thin/small structures and a downsampling branch to learn high-level discriminative knowledge. In particular, these two branches learn complementary features by residual cross-aggregation; the fusion of the complementary features from different decoding layers can be effectively accomplished through lateral connections. Meanwhile, we perform supervised prediction at all decoding layers to incorporate coarse-level features with high semantic meaning and fine-level features with high localization capability to detect multi-scale structures, especially for small/thin volumes fully. To validate the effectiveness of our S-Net, we conducted extensive experiments on the segmentation of cardiac wall and intracranial aneurysm (IA) vasculature, and quantitative and qualitative evaluations demonstrated the superior performance of our method for predicting small/thin structures in medical images.