Abstracts

Rationale: In epilepsy, 25% of patients have regular, intractable seizures, especially children with epilepsy syndromes and patients with gross brain abnormality and cognitive impairments. About half of seizures will be at night, posing problems in these vulnerable patient groups who depend on caregivers not sleeping in the same bed. A reliable seizure detection and alert system will provide a major step in patient safety, care, quality of life and disease management, however presently this is lacking. In this project a new multimodal device using an optimized combination of nonEEG sensors: audio, automated video frame analysis, ECG and 3D-accelerometry, is developed. Methods: A diagnostic study design is used to define optimal combinations of algorithms analyzing the 4 modalities in the target population: children under 18 years of age, and mentally impaired adolescents and adults with major nocturnal seizures. The multimodal device is tested in an in-hospital setting in 100 patients, simultaneously with the gold standard of clinical video-EEG. Several methods of classification and regression analysis are explored to define optimal sets of thresholds of the separate modalities for the highest possible detection rate and lowest rate of false positive outcomes. Patient factors and inevitable dependencies due to repeated measures are taken into account. Results: An interim-analysis of results of 50 patients will be presented in December 2012. Conclusions: In this study, the validity of a newly developed device for home detection of epileptic seizures during sleep in children and mentally impaired adults with major nocturnal seizures is tested.