AI Trends

Trending story

TRACER Achieves 85% Action Detection Accuracy in Test Environment

1 reports1 story updatesUpdated Nov 20 10:35

Understand this story

AI overview

On November 20, a multimodal system called TRACER, developed by a Stanford team, achieved 85% action detection accuracy in a test environment and answered physicians' queries correctly 61% of the time. The system combines continuous bedside video from the ICU with electronic health records to track changes in patient activity and help identify signs of delirium. In a cohort of 280 ICU patients, those who screened positive for delirium exhibited markedly lower levels of purposeful activity. There are currently no other evidence-supported follow-up developments.

AI-generated from public reports

Latest development · Report

Stanford Team Develops TRACER to Monitor ICU Patients Around the Clock for Early Signs of Delirium

Developments

1 confirmed developments

  1. TRACER achieves 85% action detection accuracy in a test environment

    Source:Stanford HAI News

Reports and discussion timeline

Browse reports and discussion by time. Their evidence levels remain distinct.

Nov 20

  1. Stanford HAI NewsFirst-partySelected

    Stanford Team Develops TRACER to Monitor ICU Patients Around the Clock for Early Signs of Delirium

    A Stanford team developed TRACER, a multimodal system that combines continuous bedside video from the ICU with electronic health records to track changes in patient activity and help identify signs of delirium. In a test environment, its activity detection accuracy was 85%, and it correctly answered 61% of physicians’ queries. In a cohort of 280 ICU patients, those who screened positive for delirium exhibited markedly lower levels of purposeful activity.

Related stories

  1. Related storiesOpenAI launches GPT-6.1 Terra on ChatGPT Work and Codex
  2. Related storiesTsinghua and GigaAI propose the State World Models method of settling rules first and then rendering
  3. Related storiesGoogle Research releases a multi-agent framework for AI video co-directors
  4. Related storiesEnglish translation is being prepared.