Dynamic evolution of brain structural patterns in liver transplantation recipients: a longitudinal study based on 3D...

Area of Science:

• Neuroimaging
• Medical Science
• Deep Learning Applications

Background:

• Liver transplantation (LT) recipients experience complex physiological changes.
• Monitoring long-term brain health after LT is crucial for patient outcomes.
• Neuroanatomic biomarkers can track subtle brain structural alterations.

Purpose of the Study:

• To investigate the dynamic changes in brain structure among liver transplant recipients.
• To utilize a deep learning model to quantify longitudinal brain health evolution post-LT.
• To identify specific brain networks associated with post-transplant changes.

Main Methods:

• A 3D Convolutional Neural Network (3D-CNN) model was trained on T1-weighted MRI scans from 3609 healthy individuals.
• The model predicted brain age difference (PAD) in 60 LT recipients and 134 controls at baseline and 1, 3, and 6 months post-surgery.
• Network occlusion sensitivity analysis identified key brain networks influencing age prediction.

Main Results:

• LT recipients showed a significant increase in PAD at baseline (+5.74 years) and 1 month post-surgery (+9.18 years).
• Brain age gradually decreased after 1 month but remained higher than chronological age.
• Patients with a history of OHE (ostium eticuli) exhibited higher PAD, particularly at 1 month post-LT.

Conclusions:

• Liver transplant recipients exhibit an inverted U-shaped dynamic change in brain structural patterns within the early post-transplant period.
• Brain aging initially worsens post-LT, with primary sensory networks playing a key role in these alterations.
• The findings highlight the importance of monitoring neuroanatomic changes and their contributing networks after liver transplantation.