Zhaolin Chen

This study evaluates filtering techniques for fPET data, finding that dNLM and MRI-MRF filters improve signal-to-noise ratio and test-retest reliability, with dNLM reducing sample size needs.

This study uses a Network Diffusion Model to predict neurodegeneration patterns in limb-onset ALS by correlating simulated pathology spread with MRI-measured atrophy.

The paper introduces an atlas-free method using a deep recurrent auto-encoder to parcellate tractograms by learning latent streamline embeddings for unsupervised clustering and sub-bundle analysis.

The paper introduces an open dataset comparing three FDG-fPET radiotracer administration protocols using simultaneous BOLD-fMRI scanning.

This paper improves the temporal resolution of functional PET by integrating anatomical MRI knowledge to reduce noise and partial volume errors in brain metabolism mapping.

This study uses simultaneous high-temporal resolution FDG-fPET and fMRI to compare metabolic and haemodynamic connectivity in the human brain, finding that metabolic connectivity is non-ergodic and differs from static FDG-PET covariance.

This study uses QSM to assess iron dysregulation in the motor cortex of limb-onset ALS patients over six months, finding increased QSM in lumbar-onset patients compared to controls and no significant change over time.

This study compares independent component analysis (ICA) and general linear modelling (GLM) for analyzing task and resting state fPET data, showing both methods can identify task-related brain activation.

This paper introduces Arcana, a framework for creating adaptable neuroimaging workflows that can be reused across different computing environments and data sources.

The paper compares continuous infusion and bolus plus infusion radiotracer protocols for PET to improve temporal resolution, achieving 16-second glucose response mapping in the human brain.

This study combines simultaneous BOLD-fMRI and FDG-PET to measure task-evoked neuronal metabolism in the human visual cortex with high temporal resolution.

This paper introduces an automated PET motion correction method using multi-contrast MR co-registration, improving PET image quality in MR-PET scans.