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

Search research articles

Related Experiment Video

Updated: Jan 16, 2026

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

VAE deep learning model with domain adaptation, transfer learning and harmonization for diagnostic classification

Gopikrishna Deshpande^1,2,3,4,5,6, Bonian Lu¹, Nguyen Huynh¹

¹Department of Electrical and Computer Engineering, Auburn University Neuroimaging Center, Auburn University, Auburn, AL, United States.

Frontiers in Neuroinformatics

|September 29, 2025

Summary

Related Concept Videos

You might also read

Related Articles

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

Sort by

Same author

Practice Patterns and Outcomes of Preoperative Mechanical Bowel Preparation and Oral Antibiotics for Restorative Elective Colorectal Surgery: Nationwide Analysis of the ACS-NSQIP Australian Data.

ANZ journal of surgery·2026

Same author

The Deep Learning Revolution in Neuroimaging: Insights from a Bibliometric Analysis (2014-2024).

Neuroinformatics·2026

Same author

Neptune: a toolbox for spinal cord functional MRI data processing and quality assurance.

bioRxiv : the preprint server for biology·2026

Same author

Non-invasive Electrophysiological Characterization of Distinctive Meditative States in a Yogi during <i>Samaadhi</i>.

International journal of yoga·2026

Same author

Topology Assisted Clustering of Temporal fMRI Brain Networks With Use-Case in Mitigating Non-Neural Multi-Site Variability.

IEEE access : practical innovations, open solutions·2026

Same author

Effects of seasonal factors on brain function: Systematic review and future perspectives.

iScience·2025

Domain adaptation using VAE-MMD improves machine learning for neurodevelopmental conditions in fMRI data. Transfer learning with additional healthy control data further enhances classification accuracy, outperforming traditional methods.

Area of Science:

Neuroimaging
Machine Learning
Computational Neuroscience

Background:

Multi-site fMRI datasets exhibit variability (sample, acquisition, scanner) obscuring neural signals.
This variability hinders accurate machine learning classification of neurodevelopmental conditions.
Domain adaptation offers a solution by aligning data distributions between source and target domains.

Purpose of the Study:

To demonstrate the utility of domain adaptation for multi-site fMRI data analysis.
To develop and evaluate a VAE-MMD deep learning model for classifying Autism, Asperger's syndrome, and controls.
To compare domain adaptation performance against statistical harmonization techniques.

Main Methods:

Developed a variational autoencoder-maximum mean discrepancy (VAE-MMD) deep learning model.

Keywords:

Autism Spectrum Disorders domain adaptation functional connectivity machine learning prediction variational autoencoder

More Related Videos

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data

Published on: June 26, 2013

Related Experiment Videos

Last Updated: Jan 16, 2026

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data

Published on: June 26, 2013

Utilized ABIDE-I as the source domain and ABIDE-II as the target domain for classification.

Incorporated transfer learning by augmenting the source domain with data from HBN and AOMIC datasets.

Main Results:

Domain adaptation from ABIDE-I to ABIDE-II significantly improved classification accuracy on ABIDE-II.
Transfer learning with additional healthy control data further boosted classification performance.
VAE-MMD achieved comparable performance to ComBat, with TL enhancing accuracy beyond statistical methods.

Conclusions:

Domain adaptation is effective for improving diagnostic classification in multi-site fMRI data.
Transfer learning with diverse healthy control data enhances model robustness and accuracy.
Publicly available datasets and models encourage further research in neuroimaging AI.