Journal DOI 10.68956/dyh5-kwte

Creative Commons source reading · Traumatic brain injury recovery

Traumatic brain injury recovery prediction by harmonizing real brain CT and synthetic brain MRI: a pilot study.

Yiming Che; Amogh Manoj Joshi; Jay Shah; Md Mahfuzur Rahman Siddiquee; Catherine D Chong; Simona Nikolova; Gina Dumkrieger; Baoxin Li; Teresa Wu; Todd J Schwedt

Originally published in Brain communications · 2026-04-06

From the article

Selected source text. JournalOS removes markup and shortens excerpts where indicated. The full article provides the complete context.

Abstract

To enhance the prediction of traumatic brain injury (mTBI) outcomes, we propose a deep learning approach that integrates brain computed tomography (CT) scans with corresponding synthetic T1-weighted magnetic resonance imaging (T1-MRI). Our method significantly outperforms the prediction using CT scans alone. TRACK-TBI Pilot dataset, which includes imaging and clinical outcome data from patients with TBI, is studied. The hypothesis is brain CT and T1-MRI complement each other and together will improve TBI prognosis compared to using either CT or T1-MRI alone. Since CT and T1-MRI may not be available for the same individual, we employed a specialized version of a generative adversarial network (GAN), known as fixed-point GAN (FP-GAN). FP-GAN was trained using unpaired CT and T1-MRI scans to generate synthetic T1-MRIs from real CT scans. This process produced pseudo-paired CT-MRI data, which was then used to train a deep learning classifier for outcome prediction. The classifier consists of dual parallel 3D ResNet-18 models, each independently processing T1-MRI and CT scans. We used Glasgow Outcome Scale-Extended (GOSE) scores at 3 months post-TBI as the measure of patient outcomes.

Discussion

The main finding of this study is that the prediction of patient outcomes 3 months following TBI is facilitated by using a DL-based harmonization model that integrates synthetic brain MRI with real brain CT. This method could improve the ability and accuracy to predict patient outcomes in the clinical setting since brain CT is more commonly available than brain MRI. Predicting patient outcomes following TBI can help determine which patients should receive early intervention, the frequency and intensity of follow-up evaluations and the need to refer patients to clinicians with special expertise in TBI.

Conclusion

In this study, we employed the FP-GAN model for harmonizing CT and T1-MRI data. We utilized CT scans along with their corresponding synthesized T1-MRI for predicting patient outcomes at 3 months following TBI. The prediction results demonstrate that incorporating synthetic MRI significantly enhances performance compared to using CT alone.

Editorial reading guide

  • How did the authors define recovery, and when did they measure it?
  • Does the outcome reflect a laboratory measure, a functional change or participation in daily life?