Semi-automated annotation of signal events in clinical EEG data

Abstract

To be effective, state of the art machine learning technology needs large amounts of annotated data. There are numerous compelling applications in healthcare that can benefit from high performance automated decision support systems provided by deep learning technology, but they lack the comprehensive data resources required to apply sophisticated machine learning models. Further, for economic reasons, it is very difficult to justify the creation of large annotated corpora for these applications. Hence, automated annotation techniques become increasingly important. In this study, we investigated the effectiveness of using an active learning algorithm to automatically annotate a large EEG corpus. The algorithm is designed to annotate six types of EEG events. Two model training schemes, namely threshold-based and volume-based, are evaluated. In the threshold-based scheme the threshold of confidence scores is optimized in the initial training iteration, whereas for the volume-based scheme only a certain amount of data is preserved after each iteration. Recognition performance is improved 2% absolute and the system is capable of automatically annotating previously unlabeled data. Given that the interpretation of clinical EEG data is an exceedingly difficult task, this study provides some evidence that the proposed method is a viable alternative to expensive manual annotation.

Deep learning-based electroencephalography analysis: a systematic review

“…Reactive ERP [17,31,211,230] Heard speech decoding [125] RSVP [31, 71, [135] [192] [4, 85,128,190] Event annotation [224] Prediction [141,200] [199]…”

Improved Manual Annotation of EEG Signals through Convolutional Neural Network Guidance

“…As more data get curated by human experts, we highlight the feasibility to iteratively improving our model through active learning. Similar work was done by Yang and colleagues ( S Yang et al, 2017 ) where the authors used self-training to improve detection performance in clinical EEGs. They did initial training on a small set of labeled data and used the model to automatically annotate unlabeled events with high-confidence scores to include them in the next training iteration, repeating the last two steps until all unlabeled data got annotated.…”