DNMDR: Dynamic Networks and Multi-view Drug Representations for Safe Medication Recommendation

TL;DR

This work tackles safe medication recommendation by addressing irregular, evolving EHR data and DDIs. It introduces DNMDR, a model that combines weighted dynamic heterogeneous networks (capturing temporal and structural patient information) with multi-view drug representations (internal molecular graphs, interactive co-occurrence and DDI graphs, and temporal history retrieval). Through end-to-end training with a composite loss, the approach achieves superior accuracy (PRAUC, Jaccard, F1) and safer drug combinations (lower DDI rate) on real-world datasets, with ablation and case studies validating the contributions of each component. The framework promises practical impact by enabling more reliable, personalized, and safe MR in clinical settings, while offering avenues for integrating external knowledge and advanced language models for further improvements.

Abstract

Medication Recommendation (MR) is a promising research topic which booms diverse applications in the healthcare and clinical domains. However, existing methods mainly rely on sequential modeling and static graphs for representation learning, which ignore the dynamic correlations in diverse medical events of a patient's temporal visits, leading to insufficient global structural exploration on nodes. Additionally, mitigating drug-drug interactions (DDIs) is another issue determining the utility of the MR systems. To address the challenges mentioned above, this paper proposes a novel MR method with the integration of dynamic networks and multi-view drug representations (DNMDR). Specifically, weighted snapshot sequences for dynamic heterogeneous networks are constructed based on discrete visits in temporal EHRs, and all the dynamic networks are jointly trained to gain both structural correlations in diverse medical events and temporal dependency in historical health conditions, for achieving comprehensive patient representations with both semantic features and structural relationships. Moreover, combining the drug co-occurrences and adverse drug-drug interactions (DDIs) in internal view of drug molecule structure and interactive view of drug pairs, the safe drug representations are available to obtain high-quality medication combination recommendation. Finally, extensive experiments on real world datasets are conducted for performance evaluation, and the experimental results demonstrate that the proposed DNMDR method outperforms the state-of-the-art baseline models with a large margin on various metrics such as PRAUC, Jaccard, DDI rates and so on.

Figures (9)

• Figure 1: A dynamic network for a single patient.
• Figure 2: The architecture of DNMDR
• Figure 3: Constructing a conditional probability matrix for the patient in a visit
• Figure 4: The structure of a EvolveGAT
• Figure 5: Performance variation with the epochs in comparison experiments.