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 Concept Videos

Imaging Studies for Cardiovascular System III: X-Ray01:20

Imaging Studies for Cardiovascular System III: X-Ray

The most common cardiovascular diagnostic test is an X-ray. It produces images of the heart, blood vessels, and adjacent structures.
Definition and Purpose
An X-ray, or radiograph, is a non-invasive method that uses ionizing radiation to take images of internal structures. It is mainly used in cardiac imaging to examine the heart, lungs, and major blood vessels, aiming to identify abnormalities in the heart's size, shape, and position, such as heart failure, congenital defects, and vascular...
X-ray Imaging01:24

X-ray Imaging

German physicist Wilhelm Röntgen (1845–1923) was experimenting with electrical current when he discovered that a mysterious and invisible "ray" would pass through his flesh but leave an outline of his bones on a screen coated with a metal compound. In 1895, Röntgen made the first durable record of the internal parts of a living human: an "X-ray" image (as it came to be called) of his wife’s hand. Scientists worldwide quickly began their own experiments with X-rays, and by 1900, X-ray was widely...
Radiological Investigation I: X-ray and CT01:30

Radiological Investigation I: X-ray and CT

Radiological investigations, including X-rays and computed tomography (CT) scans, are critical for diagnosing and evaluating various medical conditions. These imaging techniques provide valuable insights into the body's internal structures, aiding in the detection of abnormalities, assessment of disease progression, and development of treatment strategies. This article delves into two primary radiological investigations, chest X-rays and CT scans, outlining their purpose, procedures, and the...
Ultrasonography01:17

Ultrasonography

Ultrasonography is an imaging technique that uses high-frequency sound waves to visualize the body's internal structures. It is a non-invasive and safe procedure that does not involve the use of ionizing radiation, making it widely used in various medical fields. Ultrasonography is used to study heart function, blood flow in the neck or extremities, certain conditions such as gallbladder disease, and fetal growth and development.
During an ultrasonography procedure, a handheld device called a...

You might also read

Related Articles

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

Sort by
Same author

Radiomics in musculoskeletal imaging: what is its role in the era of deep learning?

Skeletal radiology·2026
Same author

Explainable Artificial Intelligence (AI) for Medical Imaging: A Framework for Bridging the AI Trust Gap.

AJR. American journal of roentgenology·2026
Same author

Portrayal of Medical Students in Artificial Intelligence- Generated Images.

WMJ : official publication of the State Medical Society of Wisconsin·2026
Same author

The impact of experience on radiology report uncertainty for early-career musculoskeletal radiologists.

Skeletal radiology·2026
Same author

Enhancing automated fracture detection in paediatric wrist X-rays with paired and unpaired cast suppression methods.

International journal of computer assisted radiology and surgery·2026
Same author

Test-Driving Radiology Artificial Intelligence Products Should Be Standard Practice.

AJR. American journal of roentgenology·2026

Related Experiment Video

Updated: Jul 15, 2026

Tree Core Analysis with X-ray Computed Tomography
06:56

Tree Core Analysis with X-ray Computed Tomography

Published on: September 22, 2023

Sharpness matters: Higher image resolution improves generalization and explainability in chest X-ray classification.

Aditya V Kulkarni1, John R Zech2, Paul H Yi3

  • 1Department of Radiology, St. Jude Children's Research Hospital, 262 Danny Thomas Pl, Memphis, TN, 38105, USA.

Scientific Reports
|July 13, 2026
PubMed
Summary

Training deep learning models on higher resolution chest X-rays (CXRs) significantly improves AI generalizability and explainability. Higher resolution images enhance model performance on external datasets and provide clearer visual explanations for diagnostic predictions.

Keywords:
Convolutional neural networks (CNN )Deep learningExplainable AIMachine learningNeural networks

Related Experiment Videos

Last Updated: Jul 15, 2026

Tree Core Analysis with X-ray Computed Tomography
06:56

Tree Core Analysis with X-ray Computed Tomography

Published on: September 22, 2023

Area of Science:

  • Artificial Intelligence in Medical Imaging
  • Radiology AI Development
  • Deep Learning for Healthcare

Background:

  • Deep learning (DL) models for chest X-rays (CXRs) are often trained on downsampled images to reduce computational cost.
  • Clinical workflows utilize high-resolution images, creating a discrepancy with AI training practices.
  • The impact of training resolution on AI explainability and generalizability in radiology remains underexplored.

Purpose of the Study:

  • To evaluate how training image resolution affects the performance, generalizability, and explainability of deep learning models for CXR classification.
  • To compare internal versus external testing performance and explanation quality across different training resolutions.
  • To provide practical guidance for designing trustworthy AI in radiology.

Main Methods:

  • Retrospective study training Convolutional Neural Networks (CNNs) on SIIM-ACR Pneumothorax and RSNA Pneumonia datasets.
  • Models were trained at six resolutions (64x64 to 1024x1024) using five-fold cross-validation.
  • Evaluated models on internal and external test sets, analyzing classification accuracy, generalizability (internal-to-external performance drop), and explanation quality (saliency map coverage and precision).

Main Results:

  • Internal performance was high across all resolutions (AUROC >0.85).
  • External generalizability significantly worsened at lower training resolutions (drops >20%), while higher resolutions (512x512 to 1024x1024) showed better consistency (4.2%-10.7% drops).
  • Higher resolutions led to more concise explanations (saliency map coverage <4% at 1024x1024) and improved external explanation quality.

Conclusions:

  • Training deep learning models at higher resolutions enhances both the generalizability and explainability of CXR classification AI.
  • Higher resolution training is crucial for developing robust and trustworthy AI systems in radiology, moving beyond solely internal performance metrics.
  • Findings offer practical guidance for optimizing AI design in medical imaging, emphasizing the importance of input image resolution.