Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Pneumothorax-II

Pneumothorax-II

Pneumothorax is a medical condition defined by the buildup of air in the pleural space between the lungs and the chest wall. This accumulation of air can lead to partial or complete lung collapse, resulting in a range of clinical manifestations. Understanding the clinical presentation and effective management strategies is crucial for healthcare professionals in providing timely and appropriate care to individuals with pneumothorax.
Clinical Manifestations:

You might also read

Related Articles

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

Sort by

Same author

Hope for OTHERS (Our Tissue Helping Enhance Research & Science): research results from the University of Pittsburgh rapid autopsy program for breast cancer.

Breast cancer research : BCR·2025

Same author

Hope for Others: Research Results from the University of Pittsburgh Rapid Autopsy Program for Breast Cancer.

bioRxiv : the preprint server for biology·2024

Same author

Anatomy-specific Progression Classification in Chest Radiographs via Weakly Supervised Learning.

Radiology. Artificial intelligence·2024

Same author

Anatomy-Guided Weakly-Supervised Abnormality Localization in Chest X-rays.

Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention·2024

Same author

Human Not in the Loop: Objective Sample Difficulty Measures for Curriculum Learning.

Proceedings. IEEE International Symposium on Biomedical Imaging·2023

Same author

Knowledge-Guided Multiview Deep Curriculum Learning for Elbow Fracture Classification.

Machine learning in medical imaging. MLMI (Workshop)·2023

Search research articles

Related Experiment Video

Updated: Dec 30, 2025

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

Retraining an open-source pneumothorax detecting machine learning algorithm for improved performance to medical

Gene Kitamura¹, Christopher Deible¹

¹University of Pittsburgh Medical Center (UPMC) and University of Pittsburgh, UPMC Department of Radiology, 200 Lothrop St., Pittsburgh, PA 15213, USA.

Clinical Imaging

|January 19, 2020

Summary

Machine learning models trained on open-source data require retraining for optimal performance on institutional chest X-ray datasets. This validation and optimization process is crucial for accurate pneumothorax detection in real-world clinical settings.

Keywords:

Artificial intelligence Chest X-ray Machine learning Neural network Pneumothorax

More Related Videos

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Related Experiment Videos

Last Updated: Dec 30, 2025

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Area of Science:

Artificial Intelligence in Medical Imaging
Machine Learning for Diagnostic Support
Radiology and Medical Image Analysis

Background:

Machine learning models trained on open-source datasets may not generalize effectively to local clinical data.
Validation and retraining are essential steps to adapt AI models for specific healthcare environments.
Pneumothorax detection on chest X-rays is a critical diagnostic task where AI can offer support.

Purpose of the Study:

To validate a machine learning model initially trained on an open-source dataset.
To optimize the model's performance for detecting pneumothorax on institutional chest X-rays.
To assess the impact of retraining on model accuracy for real-world clinical data.

Main Methods:

Retrospective study utilizing the CXR8 open-source dataset and institutional chest X-ray data.
Development of a limited supervision machine learning model incorporating localized and unlocalized pathology.
Training and validation of the model using dichotomized pneumothorax (PTX) and non-pneumothorax (non-PTX) cases from both datasets.
Performance evaluation using receiver operator curve (ROC) analysis, calculating area under the curve (AUC), sensitivity, and specificity.

Main Results:

Initial model training on CXR8 yielded an AUC of 0.90 for pneumothorax detection.
Direct inference on the institutional validation set resulted in a significantly lower AUC of 0.59.
Retraining the model with institutional data improved performance, achieving an AUC of 0.90 on the validation set.

Conclusions:

Machine learning models trained on open-source datasets often require a retraining phase to achieve optimal performance on local, real-world clinical data.
The study highlights the necessity of dataset-specific optimization for reliable AI-driven diagnostic tools in healthcare.
Retraining is a critical step to bridge the gap between open-source model performance and clinical utility.