Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jan 19, 2026

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

8.6K

Deep-learning model for predicting 30-day postoperative mortality.

Bradley A Fritz1, Zhicheng Cui2, Muhan Zhang2

  • 1Department of Anesthesiology, Washington University in St Louis, St Louis, MO, USA.

British Journal of Anaesthesia
|September 28, 2019
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Patients' Perspectives and Experiences of Participating in Anesthesia Research: A Qualitative Study.

Anesthesia and analgesia·2026
Same author

Integrating Mental Health Into Surgical Care: A Qualitative Study of a Perioperative Mental Health Intervention.

Annals of surgery open : perspectives of surgical history, education, and clinical approaches·2026
Same author

Machine learning models in anaesthesiology: bridging the gap from model training to implementation.

British journal of anaesthesia·2026
Same author

Feasibility Pilot Trial for the Trajectories of Recovery after Intravenous Propofol versus Inhaled Volatile Anesthesia (THRIVE) Pragmatic Randomized Clinical Trial.

Anesthesiology·2026
Same author

Exploring clinician and patient perspectives on choice of general anaesthetic. A report of two U.S. surveys: an anesthesia clinician Multicenter Perioperative Outcomes Group survey and a patient Amazon Mechanical Turk survey.

BJA open·2026
Same author

Cassandra's Curse.

Anesthesiology·2026
Same journal

Uncoupling immobility and unconsciousness.

British journal of anaesthesia·2026
Same journal

Single-dose intraoperative methadone QTc effects. Comment on Br J Anaesth 2026.

British journal of anaesthesia·2026
Same journal

Direct inhibition of cystathionine-β-synthase by isoflurane contributes to delayed neurocognitive recovery after isoflurane general anaesthesia in mice: interpreting reduced freezing after isoflurane. Comment on Br J Anaesth 2025; 135: 360-74.

British journal of anaesthesia·2026
Same journal

Anticholinergic syndrome or magic: Circe's use of anaesthetic alkaloids in Homer's Odyssey.

British journal of anaesthesia·2026
Same journal

Audit of procedural sedation complications using a novel digital application.

British journal of anaesthesia·2026
Same journal

Social deprivation and morbidity and mortality after surgery. Comment on Br J Anaesth 2025; 135: 1193-1202.

British journal of anaesthesia·2026
See all related articles

A new deep-learning model accurately predicts postoperative mortality by analyzing real-time physiological data. This advanced approach outperforms traditional methods, offering a dynamic assessment of patient risk during surgery.

Area of Science:

  • Medical Informatics
  • Machine Learning in Healthcare
  • Surgical Outcomes Research

Background:

  • Postoperative mortality affects 1-2% of major inpatient surgery patients.
  • Existing prediction tools are limited by their inability to capture dynamic intraoperative risk.
  • There is a need for improved methods to predict 30-day postoperative mortality.

Purpose of the Study:

  • To develop and evaluate a deep-learning algorithm for predicting 30-day postoperative mortality.
  • To compare the performance of a multipath convolutional neural network against traditional machine learning models and logistic regression.
  • To assess the utility of real-time intraoperative data for mortality prediction.

Main Methods:

  • A multipath convolutional neural network was constructed using patient characteristics, comorbidities, preoperative labs, and intraoperative data.
Keywords:
anaesthesiologydeep learningmachine learningpostoperative complicationsrisk predictionsurgery

Related Experiment Videos

Last Updated: Jan 19, 2026

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

8.6K
  • Data from 60 minutes prior to a random time point were utilized for patients undergoing surgery with tracheal intubation.
  • Model performance was benchmarked against deep neural network, random forest, support vector machine, and logistic regression models.

    • Main Results:

    • The multipath convolutional neural network achieved an area under the receiver operating characteristic curve of 0.867 (95% CI: 0.835-0.899).
    • This performance was superior to deep neural network (0.825), random forest (0.848), support vector machine (0.836), and logistic regression (0.837).
    • The study included 95,907 patients, with 1% (941) experiencing 30-day mortality.

    Conclusions:

    • A deep-learning time-series model significantly improves postoperative mortality prediction compared to models using simple data summaries.
    • The developed model can dynamically detect changes in patient risk in real time.
    • This offers a promising tool for enhanced patient monitoring and risk stratification in surgical settings.