Deep patient similarity learning for personalized healthcare

Qiuling Suo, Fenglong Ma, Ye Yuan, Mengdi Huai, Weida Zhong, Jing Gao, Aidong Zhang

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Predicting patients' risk of developing certain diseases is an important research topic in healthcare. Accurately identifying and ranking the similarity among patients based on their historical records is a key step in personalized healthcare. The electric health records (EHRs), which are irregularly sampled and have varied patient visit lengths, cannot be directly used to measure patient similarity due to the lack of an appropriate representation. Moreover, there needs an effective approach to measure patient similarity on EHRs. In this paper, we propose two novel deep similarity learning frameworks which simultaneously learn patient representations and measure pairwise similarity. We use a convolutional neural network (CNN) to capture local important information in EHRs and then feed the learned representation into triplet loss or softmax cross entropy loss. After training, we can obtain pairwise distances and similarity scores. Utilizing the similarity information, we then perform disease predictions and patient clustering. Experimental results show that CNN can better represent the longitudinal EHR sequences, and our proposed frameworks outperform state-of-the-art distance metric learning methods.

Original languageEnglish (US)
Article number8360146
Pages (from-to)219-227
Number of pages9
JournalIEEE Transactions on Nanobioscience
Volume17
Issue number3
DOIs
StatePublished - Jul 2018

Fingerprint

Health
Learning
Delivery of Health Care
Neural networks
Entropy
Distance Education
Cluster Analysis
Research
Deep learning

All Science Journal Classification (ASJC) codes

  • Biotechnology
  • Bioengineering
  • Medicine (miscellaneous)
  • Biomedical Engineering
  • Pharmaceutical Science
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Suo, Q., Ma, F., Yuan, Y., Huai, M., Zhong, W., Gao, J., & Zhang, A. (2018). Deep patient similarity learning for personalized healthcare. IEEE Transactions on Nanobioscience, 17(3), 219-227. [8360146]. https://doi.org/10.1109/TNB.2018.2837622
Suo, Qiuling ; Ma, Fenglong ; Yuan, Ye ; Huai, Mengdi ; Zhong, Weida ; Gao, Jing ; Zhang, Aidong. / Deep patient similarity learning for personalized healthcare. In: IEEE Transactions on Nanobioscience. 2018 ; Vol. 17, No. 3. pp. 219-227.
@article{e3403ebd14a14ea28e74b8c6880dd46c,
title = "Deep patient similarity learning for personalized healthcare",
abstract = "Predicting patients' risk of developing certain diseases is an important research topic in healthcare. Accurately identifying and ranking the similarity among patients based on their historical records is a key step in personalized healthcare. The electric health records (EHRs), which are irregularly sampled and have varied patient visit lengths, cannot be directly used to measure patient similarity due to the lack of an appropriate representation. Moreover, there needs an effective approach to measure patient similarity on EHRs. In this paper, we propose two novel deep similarity learning frameworks which simultaneously learn patient representations and measure pairwise similarity. We use a convolutional neural network (CNN) to capture local important information in EHRs and then feed the learned representation into triplet loss or softmax cross entropy loss. After training, we can obtain pairwise distances and similarity scores. Utilizing the similarity information, we then perform disease predictions and patient clustering. Experimental results show that CNN can better represent the longitudinal EHR sequences, and our proposed frameworks outperform state-of-the-art distance metric learning methods.",
author = "Qiuling Suo and Fenglong Ma and Ye Yuan and Mengdi Huai and Weida Zhong and Jing Gao and Aidong Zhang",
year = "2018",
month = "7",
doi = "10.1109/TNB.2018.2837622",
language = "English (US)",
volume = "17",
pages = "219--227",
journal = "IEEE Transactions on Nanobioscience",
issn = "1536-1241",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "3",

}

Suo, Q, Ma, F, Yuan, Y, Huai, M, Zhong, W, Gao, J & Zhang, A 2018, 'Deep patient similarity learning for personalized healthcare', IEEE Transactions on Nanobioscience, vol. 17, no. 3, 8360146, pp. 219-227. https://doi.org/10.1109/TNB.2018.2837622

Deep patient similarity learning for personalized healthcare. / Suo, Qiuling; Ma, Fenglong; Yuan, Ye; Huai, Mengdi; Zhong, Weida; Gao, Jing; Zhang, Aidong.

In: IEEE Transactions on Nanobioscience, Vol. 17, No. 3, 8360146, 07.2018, p. 219-227.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Deep patient similarity learning for personalized healthcare

AU - Suo, Qiuling

AU - Ma, Fenglong

AU - Yuan, Ye

AU - Huai, Mengdi

AU - Zhong, Weida

AU - Gao, Jing

AU - Zhang, Aidong

PY - 2018/7

Y1 - 2018/7

N2 - Predicting patients' risk of developing certain diseases is an important research topic in healthcare. Accurately identifying and ranking the similarity among patients based on their historical records is a key step in personalized healthcare. The electric health records (EHRs), which are irregularly sampled and have varied patient visit lengths, cannot be directly used to measure patient similarity due to the lack of an appropriate representation. Moreover, there needs an effective approach to measure patient similarity on EHRs. In this paper, we propose two novel deep similarity learning frameworks which simultaneously learn patient representations and measure pairwise similarity. We use a convolutional neural network (CNN) to capture local important information in EHRs and then feed the learned representation into triplet loss or softmax cross entropy loss. After training, we can obtain pairwise distances and similarity scores. Utilizing the similarity information, we then perform disease predictions and patient clustering. Experimental results show that CNN can better represent the longitudinal EHR sequences, and our proposed frameworks outperform state-of-the-art distance metric learning methods.

AB - Predicting patients' risk of developing certain diseases is an important research topic in healthcare. Accurately identifying and ranking the similarity among patients based on their historical records is a key step in personalized healthcare. The electric health records (EHRs), which are irregularly sampled and have varied patient visit lengths, cannot be directly used to measure patient similarity due to the lack of an appropriate representation. Moreover, there needs an effective approach to measure patient similarity on EHRs. In this paper, we propose two novel deep similarity learning frameworks which simultaneously learn patient representations and measure pairwise similarity. We use a convolutional neural network (CNN) to capture local important information in EHRs and then feed the learned representation into triplet loss or softmax cross entropy loss. After training, we can obtain pairwise distances and similarity scores. Utilizing the similarity information, we then perform disease predictions and patient clustering. Experimental results show that CNN can better represent the longitudinal EHR sequences, and our proposed frameworks outperform state-of-the-art distance metric learning methods.

UR - http://www.scopus.com/inward/record.url?scp=85046999276&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046999276&partnerID=8YFLogxK

U2 - 10.1109/TNB.2018.2837622

DO - 10.1109/TNB.2018.2837622

M3 - Article

C2 - 29994534

AN - SCOPUS:85046999276

VL - 17

SP - 219

EP - 227

JO - IEEE Transactions on Nanobioscience

JF - IEEE Transactions on Nanobioscience

SN - 1536-1241

IS - 3

M1 - 8360146

ER -