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

Properties of Fourier Transform II01:24

Properties of Fourier Transform II

285
The Fourier Transform (FT) is an essential mathematical tool in signal processing, transforming a time-domain signal into its frequency-domain representation. This transformation elucidates the relationship between time and frequency domains through several properties, each revealing unique aspects of signal behavior.
The Frequency Shifting property of Fourier Transforms highlights that a shift in the frequency domain corresponds to a phase shift in the time domain. Mathematically, if x(t) has...
285
Doppler Effect - I00:56

Doppler Effect - I

3.7K
The Doppler effect and Doppler shift were named after the Austrian physicist and mathematician Christian Johann Doppler in 1842, who conducted experiments with both moving sources and moving observers. Consider an observer standing on a street corner, observing an ambulance with a siren sound passing by at a constant speed. The observer experiences two characteristic changes in the sound of the siren. Initially, the sound increases in loudness as the ambulance approaches and decreases in...
3.7K
Temperature Dependent Deformation01:12

Temperature Dependent Deformation

180
In a nonhomogeneous rod made up of steel and brass, restrained at both ends and subjected to a temperature change, several steps are involved in calculating the stress and compressive load. Due to the problem's static indeterminacy, one end support is disconnected, allowing the rod to experience the temperature change freely. Next, an unknown force is applied at the free end, triggering deformations in the rod's steel and brass portions. These deformations are then calculated and added...
180

You might also read

Related Articles

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

Sort by
Same author

Author Correction: Geographics and bacterial networks differently shape the acquired and latent global sewage resistomes.

Nature communications·2026
Same author

Metagenomic peek into a corn mummy.

Scientific reports·2026
Same author

The role of surgical timing in survival of patients with spinal metastases.

International journal of clinical oncology·2026
Same author

Deep-learning-based breast cancer stage prediction from H&E-stained whole-slide images in resource-constrained settings.

Journal of pathology informatics·2026
Same author

A multimodal AI biomarker PATH-ORACLE improves prediction of recurrence in stage I lung adenocarcinoma.

medRxiv : the preprint server for health sciences·2026
Same author

Fractal-geometry analysis of pediatric posterior fossa tumors - a preoperative tool for prediction of histopathology.

Neurosurgical review·2026

Related Experiment Video

Updated: Aug 18, 2025

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
07:13

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities

Published on: October 27, 2023

1.2K

Physical imaging parameter variation drives domain shift.

Oz Kilim1, Alex Olar1, Tamás Joó2,3

  • 1Department of Physics of Complex Systems, Eötvös Loránd University, Budapest, Hungary.

Scientific Reports
|December 9, 2022
PubMed
Summary
This summary is machine-generated.

Domain shift in medical imaging is often caused by variations in physical imaging parameters (PIPs). Ensuring diverse PIP sampling during training improves model robustness and generalizability, reducing accuracy loss in real-world deployment.

More Related Videos

Optimized Setup and Protocol for Magnetic Domain Imaging with In Situ Hysteresis Measurement
09:43

Optimized Setup and Protocol for Magnetic Domain Imaging with In Situ Hysteresis Measurement

Published on: November 7, 2017

9.6K
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

43.0K

Related Experiment Videos

Last Updated: Aug 18, 2025

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
07:13

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities

Published on: October 27, 2023

1.2K
Optimized Setup and Protocol for Magnetic Domain Imaging with In Situ Hysteresis Measurement
09:43

Optimized Setup and Protocol for Magnetic Domain Imaging with In Situ Hysteresis Measurement

Published on: November 7, 2017

9.6K
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

43.0K

Area of Science:

  • Medical imaging analysis
  • Machine learning in healthcare
  • Statistical modeling

Background:

  • Statistical learning models assume independent and identically distributed data, which is often violated in medical imaging due to variations in human tissue, physician labeling, and physical imaging parameters (PIPs).
  • This violation leads to domain shift (DS), where deployed models exhibit significant accuracy drops when encountering new, out-of-distribution data.
  • Domain shift poses a major challenge for developing robust and dependable predictive models in medical imaging.

Purpose of the Study:

  • To provide the first empirical evidence that variations in PIPs between training and testing datasets are a significant driver of domain shift.
  • To investigate the correlation between variance in PIPs and model generalization error.
  • To isolate the effect of the physical imaging process on model generalization by controlling for other biases.

Main Methods:

  • Controlled experiments were conducted on a large mammogram dataset (44k images) from five hospitals.
  • The study controlled for population shift, prevalence shift, data selection biases, and annotation biases.
  • The sole effect of physical imaging parameter variation on model generalization was investigated using age group estimation as a proxy task.

Main Results:

  • Empirical evidence demonstrates that variations in physical imaging parameters (PIPs) significantly contribute to domain shift in medical imaging datasets.
  • A strong correlation was found between the variance in PIPs and model generalization error.
  • Covariate shift was observed due to biased sampling from a limited range of PIPs in both inter- and intra-hospital scenarios.

Conclusions:

  • Training data should be sampled evenly across the spectrum of physical imaging parameters (PIPs) to develop the most robust machine learning models.
  • Retaining and utilizing medical image generation metadata, measured in standard international units, is crucial.
  • This metadata can serve as a universal regularizing anchor, bridging distribution gaps across diverse global datasets and imaging modalities.