Background: Traumatic brain injury (TBI), which occurs commonly worldwide, is among the more costly of health and socioeconomic problems. Accurate prediction of favorable outcomes in severe TBI patients could assist with optimizing treatment procedures, predicting clinical outcomes, and result in substantial economic savings.Methods: In this study, we examined the capability of a machine learning-based model in predicting “favorable” or “unfavorable” outcomes after 6 months in severe TBI patients using only parameters measured on admission. Three models were developed using logistic regression, random forest, and support vector machines trained on parameters recorded from 2,381 severe TBI patients admitted to the neuro-intensive care unit of Rajaee (Emtiaz) Hospital (Shiraz, Iran) between 2015 and 2017. Model performance was evaluated using three indices: sensitivity, specificity, and accuracy. A ten-fold cross-validation method was used to estimate these indices.Results: Overall, the developed models showed excellent performance with AUC around 0.81, sensitivity and specificity of around 0.78. The top-three factors important in predicting 6-month post-trauma survival status in TBI patients are “Glasgow coma scale motor response,” “pupillary reactivity,” and “age.”Conclusions: Machine learning techniques might be used to predict the 6-month outcome in TBI patients using only the parameters measured on admission when the machine learning is trained using a large data set.

Three important criteria of existing convolutional neural networks (CNNs) are (1) test-set accuracy; (2) out-of-distribution accuracy; and (3) explainability. While these criteria have been studied independently, their relationship is unknown. For example, do CNNs that have a stronger out-of-distribution performance have also stronger explainability? Furthermore, most prior feature-importance studies only evaluate methods on 2-3 common vanilla ImageNet-trained CNNs, leaving it unknown how these methods generalize to CNNs of other architectures and training algorithms. Here, we perform the first, large-scale evaluation of the relations of the three criteria using 9 feature-importance methods and 12 ImageNet-trained CNNs that are of 3 training algorithms and 5 CNN architectures. We find several important insights and recommendations for ML practitioners. First, adversarially robust CNNs have a higher explainability score on gradient-based attribution methods (but not CAM-based or perturbation-based methods). Second, AdvProp models, despite being highly accurate more than both vanilla and robust models alone, are not superior in explainability. Third, among 9 feature attribution methods tested, GradCAM and RISE are consistently the best methods. Fourth, Insertion and Deletion are biased towards vanilla and robust models respectively, due to their strong correlation with the confidence score distributions of a CNN. Fifth, we did not find a single CNN to be the best in all three criteria, which interestingly suggests that CNNs are harder to interpret as they become more accurate.Preprint. Under review.

Background: Comparison of gene expression algorithms may be beneficial for obtaining disease pattern or grouping patients based on the gene expression profile. The current study aimed to investigate whether the knowledge within these data is able to group the ovarian cancer patients with similar disease pattern. Methods: Four different clustering methods were applied on 20 genes expression data of 37 women with ovarian cancer. All selected genes in this study had prominent roles in the control of the activity of the immune system, as well as the chemotaxis, angiogenesis, apoptosis, and etc. Comparison of different clustering methods such as K-means, Hierarchical, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) and Expectation-Maximization (EM) algorithm was the other aim of the present study. In addition, the percentage of correct prediction, Robustness-Performance Trade-off (RPT), and Silhouette criteria were used to evaluate the performance of clustering methods. Results: Six out of 20 genes (IFN-γ, Foxp3, IL-4, BCL-2, Oct4 and survivin) selected by the Laplacian score showed key roles in the development of ovarian cancer and their prognostic values were clinically and statistically confirmed. The results indicated proper capability of the expression pattern of these genes in grouping the patients with similar prognosis, i.e. patients alive after 5 years or dead (62.12%). Conclusion: The results revealed the better performance for k-means and hierarchical clustering methods, and confirmed the fact that by using the expression profile of these genes, patients with similar behavior can be grouped in the same cluster with acceptable accuracy level. Certainly, the useful information from these data may contribute to the prediction of prognosis in ovarian cancer patients along with other features of patients.