Hierarchical Graph Neural Network for Patient Treatment Preference Prediction with External Knowledge

Quan Li, Lingwei Chen, Yong Cai, Dinghao Wu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations


The healthcare industry has a wealth of data that can be used by researchers and medical professionals to infer a patient’s condition and intention to receive treatment using machine learning models. However, this line of research generally suffers from some limitations: (1) struggling to leverage structural interactions among patients; (2) attending to learn patient representations from electronic medical records (EMRs) but rarely considering supplementary contexts; and (3) overlooking EMR data imbalance issue. To address these limitations, in this paper, we propose a hierarchical graph neural network for patient treatment preference prediction. Doctors’ information and their viewing activities are first integrated as external knowledge with EMRs to construct the hierarchical graph, where a dual message passing paradigm is then devised to perform intra- and inter-subgraph aggregation to enrich patient representations and advance label propagation. To mitigate patient data imbalance issue, a community detection method is further designed to better prediction. Our experimental results demonstrate the state-of-the-art performance on patient treatment preference prediction.

Original languageEnglish (US)
Title of host publicationAdvances in Knowledge Discovery and Data Mining - 27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023, Proceedings
EditorsHisashi Kashima, Tsuyoshi Ide, Wen-Chih Peng
PublisherSpringer Science and Business Media Deutschland GmbH
Number of pages12
ISBN (Print)9783031333798
StatePublished - 2023
Event27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023 - Osaka, Japan
Duration: May 25 2023May 28 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13937 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science

Cite this