EEG Seizure Detection via 2D Frame Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting epileptic seizures from EEG data are inefficient and lack effective real-time response mechanisms.
Innovation Solution
A system that generates two-dimensional frames from EEG data, combining measurements and calculated values, which are then classified using a machine learning model to determine seizure occurrence and trigger corrective actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If traditional EEG seizure detection methods are used, then detection accuracy may be maintained, but detection speed and real-time response capability are insufficient
Solution Approach 1:
The patent transforms one-dimensional EEG time series data into two-dimensional frames by organizing sensor measurements and calculated values (such as amplitude, frequency, and other features) into a matrix structure with multiple rows and columns. This dimensional transformation enables the application of 2D convolutional neural networks which can simultaneously process spatial and temporal patterns, thereby improving both detection speed and accuracy without compromising measurement precision.
2Measurement precision
If complex EEG processing algorithms are applied, then seizure detection accuracy improves, but computational complexity and processing time increase
Solution Approach 1:
The patent segments the EEG processing task into distinct components: (1) generating two-dimensional frames from raw EEG data by organizing sensor measurements and calculated values, (2) applying a pre-trained 2D convolutional neural network classifier to the structured frames, and (3) interpreting classifier outputs to detect seizures. This segmentation allows each component to be optimized independently, reducing overall system complexity while maintaining high detection accuracy.
3Loss of time
If real-time seizure detection is implemented, then patient safety and intervention timing improve, but system reliability and false alarm rates may be affected
Solution Approach 1:
The patent employs a pre-trained 2D convolutional neural network classifier that has been trained on extensive EEG data before deployment. This preliminary training allows the system to quickly classify new EEG frames in real-time without requiring complex computations during actual seizure detection. The pre-trained model enables rapid response while maintaining high reliability by leveraging patterns learned during the training phase, thus reducing false alarms while preserving real-time detection capability.
Data Source
AI summary
Methods and systems for detecting seizures include generating two-dimensional frames that each include a first set of elements that store measurements from sensors and a second set of elements that store values calculated from said measurements. The two-dimensional frames are classified using a machine learning model. It is determined that a subject experienced a seizure during a measurement interval based on an output of the machine learning model. A corrective action is performed responsive to the determination that the subject experienced a seizure.