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

Residuals and Least-Squares Property01:11

Residuals and Least-Squares Property

9.8K
The vertical distance between the actual value of y and the estimated value of y. In other words, it measures the vertical distance between the actual data point and the predicted point on the line
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
9.8K
Difference from Background: Limit of Detection01:05

Difference from Background: Limit of Detection

8.9K
The limit of detection (LOD) is the smallest amount of analyte that can be distinguished from the background noise. The LOD value corresponds to the concentration at which the analyte signal is three times larger than the standard deviation of the blank signal. Below this value, the analyte signal cannot be differentiated from the background noise. It is calculated by dividing the calibration slope by 3 times the standard deviation of the blank signals.
The LOD indicates the presence or absence...
8.9K
Reducing Line Loss01:18

Reducing Line Loss

444
In a three-phase circuit, line loss is an indicator of energy dissipated as heat due to the resistance of transmission lines. To address this, incorporating transformers into the system—a step-up transformer at the source and a step-down transformer at the load—is a strategic solution. Two three-phase transformers are introduced to improve this.
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss in...
444
Deconvolution01:20

Deconvolution

685
Deconvolution, also known as inverse filtering, is the process of extracting the impulse response from known input and output signals. This technique is vital in scenarios where the system's characteristics are unknown, and they must be inferred from the observable signals.
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...
685
Downsampling01:20

Downsampling

771
When considering a sampled sequence with zero values between sampling instants, one can replace it by taking every N-th value of the sequence. At these integer multiples of N, the original and sampled sequences coincide. This process, known as decimation, involves extracting every N-th sample from a sequence, thereby creating a more efficient sequence.
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
771
Upsampling01:22

Upsampling

693
Managing signal sampling rates is essential in digital signal processing to maintain signal integrity. A decimated signal, characterized by a reduced frequency range due to its lower sampling rate, can be upsampled by inserting zeros between each sample. This upsampling process expands the original spectrum and introduces repeated spectral replicas at intervals dictated by the new Nyquist frequency. To refine this zero-inserted sequence, it is passed through a lowpass filter with a cutoff...
693

You might also read

Related Articles

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

Sort by
Same author

Submillimeter postmortem and in vivo diffusion and susceptibility magnetic resonance imaging to characterize cortical micro- and meso-structures.

Research square·2026
Same author

CSF turnover dysfunction: a hidden early biomarker in iRBD?

NPJ Parkinson's disease·2026
Same author

Improved Multiscale Structural Mapping with Supervertex Vision Transformer for the Detection of Alzheimer's Disease Neurodegeneration.

Human brain mapping·2026
Same author

Treatment of Clinically Diagnosed Alzheimer's Disease by External Counterpulsation A Randomized Clinical Trial.

American journal of Alzheimer's disease and other dementias·2026
Same author

Variation in systemic biological health risk is associated with white matter integrity differences in young adult veterans.

Brain imaging and behavior·2026
Same author

Novel neuroimaging biomarker of cortical gray/white border integrity in blast related-mTBI.

AJNR. American journal of neuroradiology·2026

Related Experiment Video

Updated: Jun 6, 2026

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Noise2Average: An iterative residual learning strategy for image denoising without clean data.

Zihan Li1, Ziyu Li2, Berkin Bilgic3,4,5

  • 1School of Biomedical Engineering, Tsinghua University, Beijing, P.R. China.

Imaging Neuroscience (Cambridge, Mass.)
|March 27, 2026
PubMed
Summary

Noise2Average is a new self-supervised deep learning method that effectively denoises Magnetic Resonance Imaging (MRI) data. It improves image quality and quantitative metrics without needing high-SNR reference scans, making advanced denoising more accessible.

Keywords:
diffusion tensor imagingmagnetic resonance imagingself-supervised learningtransfer learning

Related Experiment Videos

Last Updated: Jun 6, 2026

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Area of Science:

  • Medical Imaging
  • Neuroscience
  • Artificial Intelligence

Background:

  • Magnetic Resonance Imaging (MRI) quality is often limited by noise, hindering clinical and research applications.
  • Deep learning denoising methods show promise but typically require high-signal-to-noise ratio (SNR) reference data, limiting their practical use.
  • Conventional denoising techniques may not fully preserve image details or quantitative accuracy.

Purpose of the Study:

  • To develop a novel self-supervised deep learning strategy, Noise2Average, for denoising MRI data using multiple noisy repetitions.
  • To enhance the feasibility and accessibility of deep learning-based MRI denoising.
  • To evaluate Noise2Average's performance against existing methods in preserving image quality and quantitative metrics.

Main Methods:

  • Proposed an iterative residual learning strategy (Noise2Average) for denoising MRI data with multiple repetitions.
  • Employed transfer learning for subject-specific self-supervised training by fine-tuning a pre-trained convolutional neural network (CNN).
  • Applied the method to various MRI data types including T1-weighted and diffusion-weighted images (DWI).

Main Results:

  • Noise2Average demonstrated superior preservation of image sharpness and textural details compared to Noise2Noise, BM4D, and AONLM.
  • The method produced more accurate quantitative microstructural metrics from diffusion tensor imaging (DTI) data.
  • Denoising performance was comparable to supervised learning methods, while significantly reducing data and time requirements.

Conclusions:

  • Noise2Average offers a practical and effective solution for deep learning-based MRI denoising without requiring high-SNR reference data.
  • The strategy enhances the feasibility and accessibility of advanced denoising techniques for clinical and neuroscientific studies.
  • Noise2Average improves image quality and quantitative accuracy, potentially benefiting a wider range of MRI applications.