PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 16, 2026Frontiers in Neuroinformatics2 citationsOpen Access

CycleGAN Models For Brain MRI Synthesis In Multiple Sclerosis

CycleGAN models show consistent brain MRI synthesis across datasets supporting downstream tissue characterization in multiple sclerosis

View Full Paper

Authors

SSShayan ShahrokhiOOOlayinka OladosuRTRehman Tariq

Discussion

Loading...

Member takes

Overview

Demonstrates reliable brain MRI synthesis for tissue characterization in multiple sclerosis, suggesting its clinical utility.

Key Points

  • This research aims to evaluate the effectiveness of CycleGAN models for synthesizing brain MRI images for multiple sclerosis diagnosis.
  • Compared CycleGAN with Pix2Pix for T1 and T2-weighted brain MRI synthesis.
  • Utilized datasets from HCP with 1,113 healthy participants and PPMI with 318 participants.
  • Evaluated synthesized images for lesion detection, brain volumetry, and lesion texture analysis.
  • All CycleGAN models performed well; however, Pix2Pix showed better results with streamlined datasets (p < 0.001).
  • CycleGAN peak signal-to-noise ratio ranged from 24.860–28.570, while Pix2Pix ranged from 28.520–31.100.
  • High similarity observed between synthesized and source images in utility tests.
Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Shahrokhi et al. (2026) studied this question.

synapsesocial.com/papers/69b79d538166e15b153aabe3https://doi.org/10.3389/fninf.2026.1762794
Ask AI
Helpful
Bookmark
Share
View Full Paper