An Experimental Study of Time Series Based Patient Similarity with Graphs

Finding similarities between patients has been used to effectively and reliably predict diagnoses and guide treatments. However, Electronic Health Records (EHRs) contain characteristics that make analysis and application difficult. Firstly, it is difficult to compare two patients’ time series. Also, EHRs contain a vast amount of data, which proves to be a significant barrier to developing efficient systems for the widespread use of patient similarity. In this paper, we introduce a novel graph representation of time series EHRs. Our method compresses a patient’s time series medical records to reduce the storage required by more than 50%. Our paper also presents similarity metrics that can be applied to vector and graph representations of patient’s time series medical records and assesses the general performance for suggested metrics.

[1]  Jin Wang,et al.  Large-Scale Frequent Episode Mining from Complex Event Sequences with Hierarchies , 2019, ACM Trans. Intell. Syst. Technol..

[2]  Jin Wang,et al.  Scalable Metric Similarity Join Using MapReduce , 2019, 2019 IEEE 35th International Conference on Data Engineering (ICDE).

[3]  Yong Zhang,et al.  Hierarchical Inter-Attention Network for Document Classification with Multi-Task Learning , 2019, IJCAI.

[4]  Jin Wang,et al.  Learn Smart with Less: Building Better Online Decision Trees with Fewer Training Examples , 2019, IJCAI.

[5]  Hui Xiong,et al.  Temporal Phenotyping from Longitudinal Electronic Health Records: A Graph Based Framework , 2015, KDD.

[6]  Yannis Papakonstantinou,et al.  Approximate Analytics System over Compressed Time Series with Tight Deterministic Error Guarantees , 2020, Proc. VLDB Endow..

[7]  Carlo Zaniolo,et al.  MF-Join: Efficient Fuzzy String Similarity Join with Multi-level Filtering , 2019, 2019 IEEE 35th International Conference on Data Engineering (ICDE).

[8]  Yong Zhang,et al.  CLMed: A Cross-lingual Knowledge Graph Framework for Cardiovascular Diseases , 2019, WISA.

[9]  Yizhou Sun,et al.  SimGNN: A Neural Network Approach to Fast Graph Similarity Computation , 2018, WSDM.

[10]  Carlo Zaniolo,et al.  Boosting approximate dictionary-based entity extraction with synonyms , 2020, Inf. Sci..

[11]  Jin Wang,et al.  Modeling Patient Visit Using Electronic Medical Records for Cost Profile Estimation , 2018, DASFAA.

[12]  Mahdi Niamanesh,et al.  ScaDiPaSi: An Effective Scalable and Distributable MapReduce-Based Method to Find Patient Similarity on Huge Healthcare Networks , 2015, Big Data Res..

[13]  Yang Liu,et al.  graph2vec: Learning Distributed Representations of Graphs , 2017, ArXiv.

[14]  Chunxiao Xing,et al.  Discovering Subsequence Patterns for Next POI Recommendation , 2020, IJCAI.

[15]  Yong Zhang,et al.  How to Empower Disease Diagnosis in a Medical Education System Using Knowledge Graph , 2019, WISA.

[16]  Philip Chan,et al.  Toward accurate dynamic time warping in linear time and space , 2007, Intell. Data Anal..

[17]  Kaspar Riesen,et al.  Approximate graph edit distance computation by means of bipartite graph matching , 2009, Image Vis. Comput..

[18]  Weitong Chen,et al.  Learning Fine-Grained Patient Similarity with Dynamic Bayesian Network Embedded RNNs , 2019, DASFAA.

[19]  Fei Wang,et al.  Measuring Patient Similarities via a Deep Architecture with Medical Concept Embedding , 2016, 2016 IEEE 16th International Conference on Data Mining (ICDM).

[20]  Yong Zhang,et al.  A Hierarchical Framework for Top-k Location-Aware Error-Tolerant Keyword Search , 2019, 2019 IEEE 35th International Conference on Data Engineering (ICDE).

[21]  Peter Szolovits,et al.  MIMIC-III, a freely accessible critical care database , 2016, Scientific Data.

[22]  Yong Zhang,et al.  Application of Patient Similarity in Smart Health: A Case Study in Medical Education , 2019, WISA.