Search for: All records

Objective: This study develops new machine learning architectures that are more adept at detecting interictal epileptiform discharges (IEDs) in scalp EEG. A comparison of results using the average precision (AP) metric is made with the proposed models on two datasets obtained from Baptist Hospital of Miami and Temple University Hospital. Methods: Applying graph neural networks (GNNs) on functional connectivity (FC) maps of different frequency sub-bands to yield a novel architecture we call FC-GNN. Attention mechanism is applied on a complete graph to let the neural network select its important edges, hence bypassing the extraction of features, a model we refer to as CA-GNN. Results: On the Baptist Hospital dataset, the results were as follows: Vanilla Self-Attention → 0.9029 ± 0.0431, Hierarchical Attention → 0.8546 ± 0.0587, Vanilla Visual Geometry Group (VGG) → 0.92 ± 0.0618, Satelight → 0.9219 ± 0.046, FC-GNN → 0.9731 ± 0.0187, and CA-GNN → 0.9788 ± 0.0125. In the same order, the results on the Temple University Hospital dataset are 0.9692, 0.9113, 0.97, 0.9575, 0.963, and 0.9879. Conclusion: Based on the good results they yield, GNNs prove to have a strong potential in detecting epileptogenic activity. Significance: This study opens the door for the discovery of the powerful role played by GNNs in capturing IEDs, which is an essential step for identifying the epileptogenic networks of the affected brain and hence improving the prospects for more accurate 3D source localization. 

This study establishes the Orientation Relationship (OR) between the austenitic and martensitic phases of the new Shape Memory Alloy (SMA) FeMnNiAl from both experiments and analytical modeling. Through Transmission Electron Microscopy (TEM) and Electron Back-Scatter Difraction, three distinct ORs, namely the Nishiyama-Wassermann (N-W), Pitsch, and Kurdjumov–Sachs (K-S) ORs are established. The observations of non-unique ORs are explained using the energy-minimization theory of martensite revealing dependence of OR on the internal morphology of the martensitic phase, whether twinned or stackingfaulted. It is shown that the twin-variants of an internally twinned martensitic structure individually explain the Pitsch and K-S ORs. The N-W OR was observed in a stackingfaulted substructure of martensite. Through a novel extension to the energy-minimization theory for stacking-faulted substructures, the N-W OR is explained. Thus, the current study challenges the notion of OR as a material-characteristic and reveals a dependence of the OR on the internal substructure of the martensitic phase in SMAs, further establishing the OR for the new SMA FeMnNiAl.