Predicting functional outcomes after an Ischemic Stroke (IS) is highly valuable for patients and desirable for physicians. This facilitates physicians to set reasonable goals for patients and cooperate with patients and relatives effectively, and furthermore to reach common after-stroke care decisions for recovery and make exercise plans to facilitate rehabilitation. The objective of this research is to apply three current Deep Learning (DL) approaches for 6-month IS outcome predictions, using the openly accessible International Stroke Trial (IST) dataset. Furthermore, another objective of this research is to compare these DL approaches with machine learning (ML) for performing in clinical prediction. After comparing various ML methods (Deep Forest, Random Forest, Support Vector Machine, etc.) with current DL frameworks (CNN, LSTM, Resnet), the results show that DL doesn’t outperform ML significantly. DL methods and reporting used for analyzing structured medical data should be developed and improved.

Ischemic stroke subtyping was not only highly valuable for effective intervention and treatment, but also important to the prognosis of ischemic stroke. The manual adjudication of disease classification was time-consuming, error-prone, and limits scaling to large datasets. In this study, an integrated machine learning approach was used to classify the subtype of ischemic stroke on The International Stroke Trial (IST) dataset. We considered the common problems of feature selection and prediction in medical datasets. Firstly, the importances of features were ranked by the Shapiro-Wilk algorithm and Pearson correlations between features were analyzed. Then, we used Recursive Feature Elimination with Cross-Validation (RFECV), which incorporated linear SVC, Random-Forest-Classifier, Extra-Trees-Classifier, AdaBoost-Classifier, and Multinomial-Naïve-Bayes-Classifier as estimator respectively, to select robust features important to ischemic stroke subtyping. Furthermore, the importances of selected features were determined by Extra-Trees-Classifier. Finally, the selected features were used by Extra-Trees-Classifier and a simple deep learning model to classify the ischemic stroke subtype on IST dataset. It was suggested that the described method could classify ischemic stroke subtype accurately. And the result showed that the machine learning approaches outperformed human professionals.

The challenge of semantic segmentation with scarce pixel-level annotations has induced many self-supervised works, however most of which essentially train an image encoder or a segmentation head that produces finer dense representations, and when performing segmentation inference they need to resort to supervised linear classifiers or traditional clustering. Segmentation by dataset-level clustering not only deviates the real-time and end-to-end inference practice, but also escalates the problem from segmenting per image to clustering all pixels at once, which results in downgraded performance. To remedy this issue, we propose a novel self-supervised semantic segmentation training and inferring paradigm where inferring is performed in an end-to-end manner. Specifically, based on our observations in probing dense representation by image-level self-supervised ViT, i.e. semantic inconsistency between patches and poor semantic quality in non-salient regions, we propose prototype-image alignment and global-local alignment with attention map constraint to train a tailored Transformer Decoder with learnable prototypes and utilize adaptive prototypes for segmentation inference per image. Extensive experiments under fully unsupervised semantic segmentation settings demonstrate the superior performance and the generalizability of our proposed method.

Authors: Tuhin Subhra Santra, Pen-Cheng Wang and Fang Gang Tseng