Publications by authors named "Sahar Sayfollahi"
This study evaluates the reproducibility of machine learning models that integrate radiomics and deep features (features extracted from a 3D autoencoder neural network) to classify various brain hemorrhages effectively. Using a dataset of 720 patients, we extracted 215 radiomics features (RFs) and 15,680 deep features (DFs) from CT brain images. With rigorous screening based on Intraclass Correlation Coefficient thresholds (>0.
View Article and Find Full Text PDFArticle Synopsis
- - The study explored the use of a deep learning model to predict COVID-19 patient outcomes based on chest CT images, aiming to improve its clinical application through deep privacy-preserving federated learning (DPFL).
- - A total of 3,055 patients from 19 medical centers were analyzed, with the data being divided for training, validation, and testing to evaluate model performance using metrics like accuracy and sensitivity.
- - The results showed that the centralized model achieved an accuracy of 76% and the DPFL model had an accuracy of 75%, with both models demonstrating similar specificity and comparable area under the curve (AUC) values, suggesting no significant statistical differences between the two approaches.