A Study of Detection of Paroxysmal Events Utilizing Computer Vision and Machine Learning (USF)
A Study of Detection of Paroxysmal Events Utilizing Computer Vision and Machine Learning (USF)
Increased computational power has made it possible to implement complex image recognition tasks and machine learning to be implemented in every day usage. The computer vision and machine learning based solution used in this project (Nelli) is an automatic seizure detection and reporting method that has a CE mark for this specific use.
The present study will provide data to expand the utility and detection capability of NELLI and enhance the accuracy and clinical utility of automated computer vision and machine learning based seizure detection.
This is a prospective, blind comparison to the clinical gold standard for seizure characterization. This study is intended to compare the Nelli Software's ability to identify seizure events to vEEG review in adults with suspected nighttime seizures. Simultaneously, Nelli will continuously record audio and video while video-electroencephalography (vEEG) is recorded per typical standard of care. Events with positive motor manifestations will be independently identified, following standard clinical practice, by three epileptologists using clinical vEEG data. Nelli Software will review the audio and video data and independently identify events with positive motor manifestations. The outcomes of event identification will be compared between Epileptologists and the Nelli Software. For each category of event captured the positive percent agreement will be calculated using the exact binomial method. The primary endpoint of this study is to demonstrate that Nelli is able to identify seizures that have a positive motor component with a sensitivity of > 70%.
Inclusion Criteria:
Exclusion Criteria:
-