AHE Detection With a Hybrid Intelligence Model in Smart Healthcare

Acute hypotensive episode (AHE) is one of the postoperative complications with the high sudden death rate in the intensive care unit (ICU), which threatens patients' life safety seriously. Therefore, it is meaningful, valuable, and essential for the analysis including the prediction of AHE before its actual occurrence. In this paper, a novel methodology combining with deep network and multiple-gene expression programming is presented to predict the AHE automatically which will improve the efficiency of prevention and lighten the burden of diagnosis for ICU doctors. The experimental data (time series data) are obtained from multi-intelligent monitoring in intensive care II (MIMICII). The sliding window (SW) is used to generate the data stream segments from the original sources. A decomposition method (ensemble empirical mode decomposition (EEMD)) for highly complex and the nonlinear signal process is used to decompose each data stream segments into a group of intrinsic mode functions (IMFs). A novel three-layer auto-encoder network is then presented for the learning and extracting of features automatically in the mean arterial pressure (MAP) time series integrated with the SW, EEMD, and deep learning (DL) techniques. At last, the feature sets are used to construct classification models using multiple gene expression programming (GEP) classifier. The result shows that the proposed method has higher prediction accuracy of 86.16% and robustness in the prediction of the AHE.

Abstract licence: CC BY-NC-ND

35 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.