Publications by authors named "A Irimia"
Article Synopsis
- Post-traumatic epilepsy (PTE) is a common and challenging condition that arises after traumatic brain injury (TBI), but predicting its occurrence is difficult with existing methods.
- This study explored using machine learning to identify imaging features like lesion volumes and resting-state fMRI measures to improve predictions of PTE.
- The kernel support vector machine (KSVM) model was found to be the most effective, achieving a prediction accuracy of 0.78 AUC, and highlighted significant differences in the brain's temporal lobes and cerebellum between PTE and non-PTE patients.
The black box nature of deep neural networks (DNNs) makes researchers and clinicians hesitant to rely on their findings. Saliency maps can enhance DNN explainability by suggesting the anatomic localization of relevant brain features. This study compares seven popular attribution-based saliency approaches to assign neuroanatomic interpretability to DNNs that estimate biological brain age (BA) from magnetic resonance imaging (MRI).
View Article and Find Full Text PDFThe black box nature of deep neural networks (DNNs) makes researchers and clinicians hesitant to rely on their findings. Saliency maps can enhance DNN explainability by suggesting the anatomic localization of relevant brain features. This study compares seven popular attribution-based saliency approaches to assign neuroanatomic interpretability to DNNs that estimate biological brain age (BA) from magnetic resonance imaging (MRI).
View Article and Find Full Text PDFThe purpose of this study is to systematically examine the basic fluid dynamics associated with a fully liquid region within a porous material. This work has come about as a result of our investigation on the ocular fluid dynamics and transport process in a partially liquefied vitreous humor. The liquid is modeled as a sphere with Stokes flow while the surrounding infinite porous region is described by Brinkman flow.
View Article and Find Full Text PDF