A multilayer approach based on image processing and computer vision for features extraction from captured images of EEG test report
EEG is one of the major medical tests to identify the brain disorders by measuring the electrical waves generated by brain activity. Over the past few decades there has been significant research in the area of deep learning to understand the generated signals and diagnose the diseases by classifying the signal waves. Representation of these signal waves in the computer and then using them in the deep learning algorithms for diagnosis, has always been a major challenge- the features. During the research era of automated EEG diagnosis & analysis (auto-EEG-DA) most technique developed for features extraction focus on the output directly generated by EEG machines which is in form of Time-Frequency based values. A standard format named European Data Format (.edf) has been defined to represent the signal values in form of value of electric signals on a continuous timeline. In the usability viewpoint of Automated Diagnosis, the said format restriction is a barrier to the true essence of usability. I have proposed a method which would add to the usability of auto-EEG-DA by extracting features from captured images of EEG test report.
Meeting ID: 929 8873 2400