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 Videos

Machine Learning Prediction of Hospital Stay in Pediatric Typhoid Intestinal Perforation: Pilot Study.

A C Chijioke1, S O Agodirin2, A A Nasir3

  • 1Senior Registrar, Pediatric Surgery, University of Ilorin Teaching Hospital, Ilorin, Kwara State, Nigeria.

The Journal of Surgical Research
|April 28, 2026
PubMed
Summary

Related Concept Videos

Esophageal Perforation-II: Clinical Manifestations and Management01:28

Esophageal Perforation-II: Clinical Manifestations and Management

1.1K
Esophageal perforations manifest in various clinical forms, influenced by factors such as the perforation's cause and location (cervical, intrathoracic, or intra-abdominal), the extent of contamination, and potential injury to adjacent mediastinal structures. The timing between the perforation occurrence and treatment initiation also affects the clinical presentation.
Clinical Manifestations:
1.1K
Esophageal Perforation-I: Introduction01:22

Esophageal Perforation-I: Introduction

844
Esophageal perforation is a severe medical condition characterized by a breach in the integrity of the esophageal wall. This breach can occur due to various factors such as trauma, medical procedures, or underlying diseases. When the esophageal wall is compromised, it allows food, fluids, and digestive juices into the chest cavity or adjacent structures, leading to potential complications and health risks.
The location of esophageal perforation can vary, occurring anywhere along the esophagus....
844

You might also read

Related Articles

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

Sort by
Same author

Underlying Medical Conditions Identified in Patients with In-Hospital and Out-of-Hospital Cardiac Arrest in Nigeria.

Nigerian journal of clinical practice·2026
Same author

Nutraceutical Effects of <i>Justicia carnea</i> Leaf Powder Supplementations on Performance, Blood Indices, Heat Shock Protein 70, Oxidative Deoxyribonucleic Acid Damage Biomarkers and Intestinal Microbes of Broiler Chickens, Under Tropical Condition.

Archives of Razi Institute·2024
Same author

Awareness and Practice of Laparoscopic Surgery among Trainee Surgeons in Nigerian Tertiary Hospitals.

West African journal of medicine·2021
Same author

Bacterial count and predisposing factors of Clostridium perfringens (targeting CPA gene) infection along with antimicrobial sensitivity in diarrheic sheep in Pakistan.

Tropical biomedicine·2021
Same author

Oxidative status and spermatozoa kinetics of rooster semen in citrus juice-based diluent.

Tropical animal health and production·2020
Same author

Treatment of colorectal cancer in Sub-Saharan Africa: Results from a prospective Nigerian hospital registry.

Journal of surgical oncology·2019

Machine learning models can predict prolonged hospital stays for pediatric typhoid intestinal perforation (TIP). The random forest model showed excellent accuracy, aiding resource allocation in limited-setting hospitals.

Area of Science:

  • Computational epidemiology
  • Pediatric surgery
  • Machine learning in healthcare

Background:

  • Typhoid intestinal perforation (TIP) significantly impacts pediatric health in resource-limited settings, leading to extended hospital stays.
  • Existing clinical decision-making lacks specific predictive models for TIP patient hospitalization duration.
  • Optimizing resource allocation and patient management requires accurate prediction of prolonged hospital stays (>8 days).

Purpose of the Study:

  • To develop and validate machine learning models for predicting prolonged hospital stay in pediatric TIP patients.
  • To compare logistic regression and random forest models for clinical decision support using readily available admission data.
  • To create a user-friendly tool for point-of-care risk stratification.

Main Methods:

Keywords:
Data augmentationLength of stayLogistic regressionMachine learningPediatric surgeryRandom forestResource-limited settingsRisk stratificationTyphoid intestinal perforation

Related Experiment Videos

  • Retrospective analysis of 90 pediatric TIP patients from a Nigerian tertiary hospital (2009-2020).
  • Evaluation of five admission parameters: age, weight, serum potassium, serum sodium, and packed cell volume.
  • Development and validation of logistic regression and random forest classifiers using dataset augmentation and cross-validation; creation of a graphical interface.

Main Results:

  • The random forest model achieved excellent discrimination with a mean Area Under the Curve (AUC) of 0.887.
  • The logistic regression model demonstrated good discrimination with a mean AUC of 0.715.
  • Serum potassium was the most significant predictor (34.1% feature importance), followed by packed cell volume and sodium.

Conclusions:

  • Machine learning models are feasible for clinical deployment in pediatric TIP risk stratification.
  • The random forest model offers superior predictive performance, while logistic regression provides high interpretability.
  • Data augmentation addressed sample size limitations; serum potassium monitoring is crucial, and external validation is recommended.