Multi-scale Value-Density Transformer with Medical Semantic Guidance for Disease Risk Prediction Based on Clinical Time Series

Jingwen Xu, Xiaoge Wei, Pong C. Yuen*

*Corresponding author for this work

Research output: Chapter in book/report/conference proceedingConference proceedingpeer-review

Abstract

Clinical time series are ubiquitous in healthcare for accurate disease risk prediction. The recent Transformer models have demonstrated superior performance in time series learning. However, these methods focus on global temporal dependency and widely utilize channel-mix or channel-independent tokenization. They ignore local dependency in clinical time series and are limited to capturing the intrinsic clinical variable correlations. To address the above issues, we present an Multi-scale Value-Density Transformer with Medical Semantic Guidance (MVMformer), which takes irregularity-aware segment-wise modeling for clinical time series and correlate diverse clinical variates based on their medical semantic affinity. Specifically, MVMformer introduces a Multi-scale Value-Density Attention to capture intra-segment characteristics in both value trends and temporal density while accommodating multi-length segments. Furthermore, MVMformer constructs a Hierarchical Medical Semantic Graph to analyze complicated variable relationships starting from detailed measurements to associated organs. Experimental results on three medical datasets demonstrate the superiority of MVMformer over existing state-of-the-art methods.

Original languageEnglish
Title of host publicationPattern Recognition
Subtitle of host publication27th International Conference, ICPR 2024, Kolkata, India, December 1–5, 2024, Proceedings, Part XXIII
EditorsApostolos Antonacopoulos, Subhasis Chaudhuri, Rama Chellappa, Cheng-Lin Liu, Saumik Bhattacharya, Umapada Pal
Place of PublicationCham
PublisherSpringer Cham
Pages163-178
Number of pages16
ISBN (Electronic)9783031783470
ISBN (Print)9783031783463
DOIs
Publication statusPublished - 1 Dec 2024
Event27th International Conference on Pattern Recognition - Kolkata, India
Duration: 1 Dec 20245 Dec 2024
https://link.springer.com/book/10.1007/978-3-031-78107-0 (Conference proceedings)
https://icpr2024.org/ (Conference website)

Publication series

NameLecture Notes in Computer Science
Volume15323
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349
NameICPR: International Conference on Pattern Recognition

Conference

Conference27th International Conference on Pattern Recognition
Abbreviated title ICPR 2024
Country/TerritoryIndia
CityKolkata
Period1/12/245/12/24
Internet address

Scopus Subject Areas

  • Theoretical Computer Science
  • General Computer Science

User-Defined Keywords

  • Clinical Time Series
  • Intensity
  • Kernelized
  • Medical Semantic
  • Segment

Fingerprint

Dive into the research topics of 'Multi-scale Value-Density Transformer with Medical Semantic Guidance for Disease Risk Prediction Based on Clinical Time Series'. Together they form a unique fingerprint.

Cite this