No abstract

Acute ischemic stroke (AIS) has been a common threat to human health and may lead to severe outcomes without proper and prompt treatment. To precisely diagnose AIS, it is of paramount importance to quantitatively evaluate the AIS lesions. By adopting a convolutional neural network (CNN), many automatic methods for ischemic stroke lesion segmentation on magnetic resonance imaging (MRI) have been proposed. However, most CNN-based methods should be trained on a large amount of fully labeled subjects, and the label annotation is a labor-intensive and time-consuming task. Therefore, in this paper, we propose to use a mixture of many weakly labeled and a few fully labeled subjects to relieve the thirst of fully labeled subjects. In particular, a multifeature map fusion network (MFMF-Network) with two branches is proposed, where hundreds of weakly labeled subjects are used to train the classification branch, and several fully labeled subjects are adopted to tune the segmentation branch. By training on 398 weakly labeled and 5 fully labeled subjects, the proposed method is able to achieve a mean dice coefficient of 0.699 ± 0.128 on a test set with 179 subjects. The lesion-wise and subject-wise metrics are also evaluated, where a lesion-wise F1 score of 0.886 and a subject-wise detection rate of 1 are achieved.

No abstract

Ischemic stroke has been a common disease in the elderly population, which can cause long-term disability and even death. However, the time window for treatment of ischemic stroke in its acute stage is very short. To fast localize and quantitively evaluate the acute ischemic stroke (AIS) lesions, many deep-learning-based lesion segmentation methods have been proposed in the literature, where a deep convolutional neural network (CNN) was trained on hundreds of fully-labeled subjects with accurate annotations of AIS lesions. Such methods, however, require a large number of subjects with pixel-by-pixel labels, making it very time-consuming in data collection and annotation. Therefore, in this paper, we propose to use a large number of weakly-labeled subjects with easy-obtained slice-level labels and a few fully-labeled ones with pixel-level annotations, and propose a semisupervised learning method. In particular, a double-path classification network (DPC-Net) was proposed and trained using the weakly-labeled subjects to detect the suspicious AIS lesions. A K-means algorithm was used on the diffusion -weighted images (DWIs) to identify the potential AIS lesions due to the a priori knowledge that the AIS lesions appear as hyperintense. Finally, a region-growing algorithm combines the outputs of the DPC-Net and the K-means to obtain the precise lesion segmentation. By using 460 weakly-labeled subjects and 5 fully-labeled subjects to train and fine-tune the proposed method, our proposed method achieves a mean dice coefficient of 0.642, and a lesion-wise F1 score of 0.822 on a clinical dataset with 150 subjects.