Loading Now

Summary of Timehr: Image-based Time Series Generation For Electronic Health Records, by Hojjat Karami et al.


TimEHR: Image-based Time Series Generation for Electronic Health Records

by Hojjat Karami, Mary-Anne Hartley, David Atienza, Anisoara Ionescu

First submitted to arxiv on: 9 Feb 2024

Categories

  • Main: Machine Learning (cs.LG)
  • Secondary: None

     Abstract of paper      PDF of paper


GrooveSquid.com Paper Summaries

GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!

Summary difficulty Written by Summary
High Paper authors High Difficulty Summary
Read the original abstract here
Medium GrooveSquid.com (original content) Medium Difficulty Summary
The proposed TimEHR model uses generative adversarial networks (GANs) to generate time series data from Electronic Health Records (EHRs). The unique challenges of EHR time series, including irregular sampling, missing values, and high dimensionality, are addressed by treating time series as images. Two conditional GANs work together: one generates missingness patterns, while the other generates time series values based on these patterns. Experimental results on three real-world EHR datasets demonstrate that TimEHR outperforms state-of-the-art methods in terms of fidelity, utility, and privacy metrics.
Low GrooveSquid.com (original content) Low Difficulty Summary
TimEHR is a new way to create fake healthcare data from real patient records. It uses special kinds of artificial intelligence called GANs. These GANs make two types of predictions: where the missing values are, and what those values should be. This helps keep patients’ information private while still being useful for doctors and researchers. The results show that TimEHR works better than other methods in making fake data that is similar to real data.

Keywords

* Artificial intelligence  * Time series