Utilizing electroencephalography (EEG) signals, this repository aims to detect, classify, and analyze seizures as well as other forms of harmful brain activity in critically ill patients.
The goal is to detect and classify seizures and other types of harmful brain activity using electroencephalography (EEG) signals recorded from critically ill hospital patients. Your work may contribute to improving electroencephalography pattern classification accuracy, with potential benefits for neurocritical care, epilepsy treatment, and drug development.
Physicians rely on EEG monitoring to detect seizures and other types of brain activity in critically ill patients. Currently, EEG analysis is predominantly manual and labor-intensive, leading to bottlenecks in diagnosis and treatment. This project aims to automate EEG analysis to accelerate diagnosis and improve treatment accuracy.
EEG segments used in the competition have been annotated by experts. Annotations include idealized patterns with high agreement, proto-patterns with mixed agreement, and edge cases where experts are split between two patterns. Detailed explanations of EEG patterns and expert annotations are provided in the dataset.
Jin Jing, Zhen Lin, Chaoqi Yang, Ashley Chow, Sohier Dane, Jimeng Sun, M. Brandon Westover. (2024). HMS - Harmful Brain Activity Classification . Kaggle. https://kaggle.com/competitions/hms-harmful-brain-activity-classification