Isotropic reconstruction of a 4-D MRI thoracic sequence using super-resolution

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Image Reconstruction

Background:

• Four-dimensional (4D) thoracic MRI effectively monitors tumor and lung anatomy.
• High temporal resolution in 4D MRI is crucial for motion artifact reduction but compromises spatial resolution.
• Current 4D MRI protocols face limitations in achieving both high temporal and spatial resolution for thoracic applications.

Purpose of the Study:

• To propose a method for reconstructing isotropic 4D MRI thoracic sequences with minimal acquisition protocol changes.
• To enhance the spatial resolution of 4D thoracic MRI for improved clinical utility.
• To advance the application of 4D MRI in thoracic radiotherapy.

Main Methods:

• A post-acquisition super-resolution technique combines three orthogonal anisotropic 4D MRI acquisitions.
• Reconstruction yields a series of isotropic volumes from anisotropic data.
• A novel phantom simulating lung tumor motion was developed for algorithm evaluation, alongside application to patient data.

Main Results:

• Super-resolution significantly enhanced image resolution and reduced partial volume effects.
• Isotropic reconstruction improved visualization and segmentation of thoracic structures in patient data.
• Both subjective and objective evaluations confirmed the effectiveness of the proposed method.

Conclusions:

• Super-resolution is an efficient method for improving 4D MRI sequence resolution.
• The technique enables the creation of isotropic 4D sequences not feasible through direct acquisition.
• Further research is needed to assess the clinical reproducibility of this method in diverse applications.