Scholarlink

Identification of patient-ventilator asynchronies must be optimized to avoid deleterious effects. Due to the difficulty of continuous evaluation of mechanical ventilator waveforms and the existence of patterns not explained in the literature, automatic identification strategies have been proposed. This work aims to present a form of unsupervised extraction of patterns of mechanical ventilation waveforms, using the self-organizing map. A total of 3051 breath cycles were divided into four clusters using two approaches for wrapping them in the same number of samples. The combination of these approaches was able to identify patterns that resamble three of the main patient-ventilator asynchronies and waveforms patterns not yet defined.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Extraction of Patterns of Mechanical Ventilation Waveforms Using Unsupervised Neural Network: A Preliminary Study

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Ricardo Gaudio de Almeida,
  • G. Motta-Ribeiro,
  • A. Giannella-Neto,
  • A. R. S. Carvalho,
  • L. M. Camilo,
  • J. Nadal

摘要

Identification of patient-ventilator asynchronies must be optimized to avoid deleterious effects. Due to the difficulty of continuous evaluation of mechanical ventilator waveforms and the existence of patterns not explained in the literature, automatic identification strategies have been proposed. This work aims to present a form of unsupervised extraction of patterns of mechanical ventilation waveforms, using the self-organizing map. A total of 3051 breath cycles were divided into four clusters using two approaches for wrapping them in the same number of samples. The combination of these approaches was able to identify patterns that resamble three of the main patient-ventilator asynchronies and waveforms patterns not yet defined.