Thomas W. Okell

This paper develops a fast few-shot transfer learning approach for reconstructing highly accelerated 4D arterial spin labeling angiography using simulated data and minimal experimental data.

The paper introduces a non-contrast MRI method that simultaneously acquires angiographic, perfusion, and structural brain images in six minutes, showing improved vessel visibility and fewer artifacts compared to conventional methods.

This paper develops a few-shot learning method for 3D TOF-MRA reconstruction, achieving high-quality results with minimal experimental data and outperforming existing methods.

This paper optimizes PCASL parameters for 7 T whole-brain perfusion imaging, achieving higher SNR efficiency and reduced RF power usage.

This paper introduces a method to achieve ultra-high temporal resolution 4D angiography using arterial spin labeling with subspace reconstruction, improving spatiotemporal fidelity and temporal resolution.

This paper develops a dynamic B0 field shimming method for 7T PCASL to improve labeling efficiency and perfusion signal by correcting off-resonance in brain-feeding arteries.

This paper introduces a subspace-based motion correction method for MRI with varying tissue contrast, improving image quality and correlation with reference images.

This paper introduces an efficient 3D cone trajectory design for improved spatial resolution and repeatability in combined angiographic and perfusion MRI using arterial spin labeling.

This paper optimizes a non-contrast MRI technique to generate 4D angiographic and perfusion images of the whole head from a single scan using radial imaging and arterial spin labeling.

This study optimizes and compares pseudo-continuous arterial spin labeling protocols for cerebral blood flow measurement, finding that a hybrid protocol provides more accurate and repeatable results.

The paper introduces SILVER, a radial sampling method for MRI that improves SNR and uniformity compared to golden ratio sampling in dynamic imaging.

This paper introduces a machine learning method for dual-calibrated fMRI to estimate OEF and CMRO2, showing reduced computation time, error, and bias compared to existing methods.

This paper demonstrates vessel-selective arterial spin labeling angiography with high acceleration and no loss of image quality compared to conventional methods.

This study demonstrates that radial k-space trajectories improve vessel-selective dynamic angiography with arterial spin labeling by enhancing SNR and reducing acquisition time.

This paper applies streamlined-qBOLD to acute stroke patients to quantify brain oxygenation parameters and their relation to tissue outcomes.

This study uses arterial spin labelling to investigate perfusion changes in patients with musculoskeletal pain during arm repositioning, finding increased perfusion in brain regions associated with pain processing.