Pediatric Seizure Prediction Model Evaluation with Hypergraph Convolutions

22 pediatric patients were used to evaluate a seizure prediction framework based on hypergraph convolutions and Kuramoto oscillator dynamics. The dataset summarizes model contributions with statistical significance and was authored by Masoud Amiri, last updated in April 2026. It is a small dataset at 5.5 KB, stored in an XLS file.

Use Cases

• Evaluate seizure prediction accuracy (e.g., 'Chimera Detection Accuracy', 'Gnn Achieved 84') against patient data from 22 pediatric subjects.
• Analyze the impact of model components like 'Layer Hypergraph Convolutions' or 'Task Prediction Heads' on performance metrics.
• Study the application of 'Kuramoto Oscillator Dynamics' and 'Delay Differential Analysis' for modeling neural interactions in epilepsy.
• Benchmark the 'Framework Transforms Multi' and 'Delta Desynchronized' features for time-series forecasting in a clinical context.

Strengths

• Model achieved a reported 84% accuracy for seizure prediction.
• Evaluation includes data from 22 pediatric patients.
• Data is shared under a permissive CC-BY-4.0 license.
• Recent update in April 2026.

Limitations

• Extremely small data scale at 5.5 KB, indicating limited raw patient data or only summary results.
• Sample size of 22 patients is small for robust statistical generalization.
• Specific data columns, row counts, and raw time-series are unknown.

Provenance

Source

figshare, author Masoud Amiri.

Collection Method

Likely results from a computational modeling study evaluating a seizure prediction framework.

Time Range

null

Freshness

Last updated April 2026.

Geography

null