Skeleton-based cerebrovascular quantitative analysis

All answers are in the images: A review of deep learning for cerebrovascular segmentation

Cerebrovascular imaging is a common examination. Its accurate cerebrovascular segmentation become an important auxiliary method for the diagnosis and treatment of cerebrovascular diseases, which has received extensive attention from researchers. Deep learning is a heuristic method that encourages researchers to derive answers from the images by driving datasets. With the continuous development of datasets and deep learning theory, it has achieved important success for cerebrovascular segmentation. Detailed survey is an important reference for researchers. To comprehensively analyze the newest cerebrovascular segmentation, we have organized and discussed researches centered on deep learning. This survey comprehensively reviews deep learning for cerebrovascular segmentation since 2015, it mainly includes sliding window based models, U-Net based models, other CNNs based models, small-sample based models, semi-supervised or unsupervised models, fusion based models, Transformer based models, and graphics based models. We organize the structures, improvement, and important parameters of these models, as well as analyze development trends and quantitative assessment. Finally, we have discussed the challenges and opportunities of possible research directions, hoping that our survey can provide researchers with convenient reference.

Automated anatomical labeling of a topologically variant abdominal arterial system via probabilistic hypergraph matching

Automated anatomical vessel labeling of the abdominal arterial system is a crucial topic in medical image processing. One reason for this is the importance of the abdominal arterial system in the human body, and another is that such labeling is necessary for the related disease diagnoses, treatments and epidemiological population analyses. We define a hypergraph representation of the abdominal arterial system as a family tree model with a probabilistic hypergraph matching framework for automated vessel labeling. Then we treat the labelling problem as the convex optimization problem and solve it with the maximum a posteriori(MAP) combined the likelihood obtained by geometric labelling with the family tree topology-based knowledge. Geometrically, we utilize XGBoost ensemble learning with an intrinsic geometric feature importance analysis for branch-level labeling. In topology, the defined family tree model of the abdominal arterial system is transferred as a Markov chain model using a constrained traversal order method and further the Markov chain model is optimized by a hidden Markov model (HMM). The probability distribution of the target branches for each candidate anatomical name is predicted and effectively embedded in the HMM model. This approach is evaluated with the leave-one-out method on 37 clinical patients' abdominal arteries, and the average accuracy is 91.94%. The obtained results are better than those of the state-of-art method with an F1 score of 93.00% and a recall of 93.00%, as the proposed method simultaneously handles the anatomical structural variability and discriminates between the symmetric branches. It is demonstrated to be suitable for labelling branches of the abdominal arterial system and can also be extended to similar tubular organ networks, such as arterial or airway networks.