UPenn and Mayo Clinic's Seizure Detection Challenge Dataset is a dataset containing epilepsy lesion annotations. It contains training and test data for humans and dogs. The training data is a 1-second EEG clip labeled "Ictal" for epileptic seizure data segments and "Interval" for non-epileptic seizure data segments. The training data is arranged sequentially, while the test data is randomly arranged.

This dataset provides Lctal training and test datasets covering the entire epileptic seizure, and provides inter-seizure data segments that cover the duration of the patient's epileptic seizure. The data comes from related challenges and was released by Kaggle in 2015.