Ensemble learning using individual neonatal data for seizure detection

TL;DR

This study proposes an ensemble learning approach for neonatal seizure detection using data from multiple institutions without data sharing, achieving accuracy comparable to centralized models through various aggregation methods.

Contribution

It introduces a novel ensemble method that combines local detectors trained on institution-specific data, addressing data sharing limitations in medical settings.

Findings

Weighted mean aggregation performed best among methods.

Ensemble accuracy approached that of a centralized model.

Dawid-Skene method marginally outperformed other schemes.

Abstract

Sharing medical data between institutions is difficult in practice due to data protection laws and official procedures within institutions. Therefore, most existing algorithms are trained on relatively small electroencephalogram (EEG) data sets which is likely to be detrimental to prediction accuracy. In this work, we simulate a case when the data can not be shared by splitting the publicly available data set into disjoint sets representing data in individual institutions. We propose to train a (local) detector in each institution and aggregate their individual predictions into one final prediction. Four aggregation schemes are compared, namely, the majority vote, the mean, the weighted mean and the Dawid-Skene method. The method was validated on an independent data set using only a subset of EEG channels. The ensemble reaches accuracy comparable to a single detector trained on all the data when sufficient amount of data is available in each institution. The weighted mean aggregation scheme showed best performance, it was only marginally outperformed by the Dawid--Skene method when local detectors approach performance of a single detector trained on all available data.