Arterial input function (AIF) is estimated from perfusion images as a basic curve for the following deconvolution process to calculate hemodynamic variables to evaluate vascular status of tissues. However, estimation of AIF is currently based on manual annotations with prior knowledge. We propose an automatic estimation of AIF in perfusion images based on a multi-stream 3D CNN, which combined spatial and temporal features together to estimate the AIF ROI. The model is trained by manual annotations. The proposed method was trained and tested with 100 cases of perfusion-weighted imaging. The result was evaluated by dice similarity coefficient, which reached 0.79. The trained model had a better performance than the traditional method. After segmentation of the AIF ROI, the AIF was calculated by the average of all voxels in the ROI. We compared the AIF result with the manual and traditional methods, and the parameters of further processing of AIF, such as time to the maximum of the tissue residue function (Tmax), relative cerebral blood flow, and mismatch volume, which are calculated in the Section Results. The result had a better performance, the average mismatch volume reached 93.32% of the manual method, while the other methods reached 85.04 and 83.04%. We have applied the method on the cloud platform, Estroke, and the local version of its software, NeuBrainCare, which can evaluate the volume of the ischemic penumbra, the volume of the infarct core, and the ratio of mismatch between perfusion and diffusion images to help make treatment decisions, when the mismatch ratio is abnormal.

Automated cerebrovascular segmentation of time-of-flight magnetic resonance angiography (TOF-MRA) images is an important technique, which can be used to diagnose abnormalities in the cerebrovascular system, such as vascular stenosis and malformation. Automated cerebrovascular segmentation can direct show the shape, direction and distribution of blood vessels. Although deep neural network (DNN)-based cerebrovascular segmentation methods have shown to yield outstanding performance, they are limited by their dependence on huge training dataset. In this paper, we propose an unsupervised cerebrovascular segmentation method of TOF-MRA images based on DNN and hidden Markov random field (HMRF) model. Our DNN-based cerebrovascular segmentation model is trained by the labeling of HMRF rather than manual annotations. The proposed method was trained and tested using 100 TOF-MRA images. The results were evaluated using the dice similarity coefficient (DSC), which reached a value of 0.79. The trained model achieved better performance than that of the traditional HMRFbased cerebrovascular segmentation method in binary pixel-classification. This paper combines the advantages of both DNN and HMRF to train the model with a not so large amount of the annotations in deep learning, which leads to a more effective cerebrovascular segmentation method.

No abstract