AMRNN: attended multi-task recurrent neural networks for dynamic illness severity prediction
- PDF / 1,087,274 Bytes
- 18 Pages / 439.642 x 666.49 pts Page_size
- 43 Downloads / 139 Views
AMRNN: attended multi-task recurrent neural networks for dynamic illness severity prediction Weitong Chen1 · Guodong Long1 · Lina Yao2 · Quan Z. Sheng3 Received: 17 March 2019 / Revised: 23 July 2019 / Accepted: 6 August 2019 / © Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract Illness severity prediction (ISP) is crucial for caregivers in the intensive care unit (ICU) while saving the life of patients. Existing ISP methods fail to provide sufficient evidence for the time-critical decision making in the dynamic changing environment. Moreover, the correlated temporal features in multivariate time-series are rarely be considered in existing machine learning-based ISP models. Therefore, in this paper, we propose a novel interpretable analysis framework which simultaneously analyses organ systems differentiated based on the pathological and physiological evidence to predict illness severity of patients in ICU. It not only timely but also intuitively reflects the critical conditions of patients for caregivers. In particular, we develop a deep interpretable learning model, namely AMRNN, which is based on the Multi-task RNNs and Attention Mechanism. Physiological features of each organ system in multivariate time series are learned by a single Long-Short Term Memory unit as a dedicated task. To utilize the functional and temporal relationships among organ systems, we use a shared LSTM task to exploit correlations between different learning tasks for further performance improvement. Real-world clinical datasets (MIMIC-III) are used for conducting extensive experiments, and our method is compared with the existing state-of-the-art methods. The experimental results demonstrated that our proposed approach outperforms those methods and suggests a promising way of evidence-based decision support. Keywords Multi-task learning · Deep learning · Illness severity prediction
This article belongs to the Topical Collection: Special Issue on Application-Driven Knowledge Acquisition Guest Editors: Xue Li, Sen Wang, and Bohan Li Weitong Chen
[email protected]
Extended author information available on the last page of the article.
World Wide Web
1 Introduction The accumulation of over 69 million EHRs from over 21.9 million (90.1% of National Participation Rate) individuals in the My Health Record System1 , has cached great attentions from machine learning and data mining communities. Learning such a large volume of data from different sources could provide strong supports for evidence-based clinical decision making in ICU, which could benefit clinical practice. However, multi-format of EHRs are abundant in terms of data categories, data types, and multivariate time series, but are usually vendor-specific and limited in scope [3]. A clinical decision in ICU is fundamentally driven by forecasting an outcome for patients in terms of quality and length of life [3]. Recently, deep learning technics have advanced the researches in ICU decision support. But, they mainly focus on mortality estimation [23] and ph
Data Loading...
