Using Deep-Learning-Based Artificial Intelligence Technique to Automatically Evaluate the Collateral Status of Multiphase CTA in Acute Ischemic Stroke

Abstract

Background: Collateral status is an important predictor for the outcome of acute ischemic stroke with large vessel occlusion. Multiphase computed-tomography angiography (mCTA) is useful to evaluate the collateral status, but visual evaluation of this examination is time-consuming. This study aims to use an artificial intelligence (AI) technique to develop an automatic AI prediction model for the collateral status of mCTA. Methods: This retrospective study enrolled subjects with acute ischemic stroke receiving endovascular thrombectomy between January 2015 and June 2020 in a tertiary referral hospital. The demographic data and images of mCTA were collected. The collateral status of all mCTA was visually evaluated. Images at the basal ganglion and supraganglion levels of mCTA were selected to produce AI models using the convolutional neural network (CNN) technique to automatically predict the collateral status of mCTA. Results: A total of 82 subjects were enrolled. There were 57 cases randomly selected for the training group and 25 cases for the validation group. In the training group, there were 40 cases with a positive collateral result (good or intermediate) and 17 cases with a negative collateral result (poor). In the validation group, there were 21 cases with a positive collateral result and 4 cases with a negative collateral result. During training for the CNN prediction model, the accuracy of the training group could reach 0.999 ± 0.015, whereas the prediction model had a performance of 0.746 ± 0.008 accuracy on the validation group. The area under the ROC curve was 0.7. Conclusions: This study suggests that the application of the AI model derived from mCTA images to automatically evaluate the collateral status is feasible.

Deep Learning and Artificial Intelligence in Action (2019–2023): A Review on Brain Stroke Detection, Diagnosis, and Intelligent Post-Stroke Rehabilitation Management

“…Table III highlights the layer types and number of layers contained in some of the extensively utilized DL models in the literature for handling brain stroke data. [26] 44 (40 3D convolution layers, 2 fully connected layers, 1 ReLu, 1 SoftMax layer) VGG-SegNet [47] 40 (2 blocks of 2 convolutional + max pooling layer, 3 blocks of 3 convolutional + max pooling layer, 3 sets of max-pooling + 3 up sampling layer, 2 blocks of max pooling + 2 up sampling layer, 3 fully connected layer, 1 SoftMax layer) VGG16 [5] 16 (2 blocks of 2 convolutional + max pooling layer, 3 blocks of 3 convolutional + max pooling layer, 1 flatten layer, 1 dense layer) 1D-CNN [10] 16 (4 blocks of 2 convolutions + 1 max-pooling layer, 1 global average poling layer, 1 dropout layer, 1 fully connected layer, 1 softmax layer) OzNet [27] 34 (7 blocks of a convolutional + a maximum pooling + a ReLU + a batch normalization layer, 2 fully connected layers, a dropout layer, a SoftMax layer, and a classification layer) ISP-Net [33] 22 (4 blocks of convolution + batch normalization + ReLu layer, 3 max-pooling layers, 5 residual blocks, 2 deconvolution layers) CNN [40] 13 (5 blocks of convolution + max-pooling layers, a flatten layer, 2 fully connected layers) AG-DCNN [67] 23 (2 convolution + max-pooling layer, 3 blocks of convolution + max-pooling layer, 3 blocks of a upsampled layer + 3 convolution + a max-pooling layer) PerfU-Net [17] 30 Gaidhani, B. R. et al, [50] used MRI-based brain stroke diagnosis utilizing CNN and DL algorithms. The suggested technique is to use semantic segmentation to identify MRI brain stroke images as abnormal or normal and to define aberrant areas.…”

Clinical Applications of Artificial Intelligence in Medical Imaging and Image Processing—A Review

“…This collection includes eight articles on detection, five on classification, four on segmentation, seven on prediction, five on quality improvement, and three on simulation. The following organs were of interest: the lungs (4), breasts (4), liver (3), brain (9), prostate (3), and others (8). Two studies were conducted on phantoms.…”