Deep learning reconstruction cuts prostate MRI acquisition time
A commercially available deep learning reconstruction (DLR) algorithm could slash the acquisition times of prostate MRIs.
New research published Tuesday in Clinical Imaging details the benefits of applying a DLR algorithm to diffusion weighted imaging (DWI) sequences to reduce exam time without sacrificing image quality. Experts involved in the study suggested that the benefits of similar tools could extend beyond just image quality, potentially increasing access to the exam as well.
“Advances in deep learning image reconstruction have emerged in recent years showing immense potential for improving prostate mpMRI image quality and reducing image noise. Protocols achieving shorter acquisition times without sacrificing image quality are highly desirable as they would presumably improve both resource utilization and patient experience,” corresponding author Mukesh G. Harisinghani, with the Department of Radiology at Massachusetts General Hospital, and colleagues noted.
For the study, researchers retrospectively reconstructed raw prostate MR image data using both conventional 2D Cartesian and DLR algorithms. The team simulated decreased DWI acquisition times by reconstructing DLR datasets containing a reduced number of excitations (NEX), and had two radiologists evaluate the resultant images for quality and the presence of image noise.
Compared to images that had been reconstructed conventionally, the radiologists reported less noise for both restricted and large field of view diffusion images that had been processed using DLR. Image quality remained consistent using DLR, while the signal-to-noise ratio saw an improvement.
Encouragingly, use of DLR reduced DWI acquisition time by 68% and 39% for restricted and large FOV acquisitions. The authors suggested that the shortened exam time would presumably allow departments to accommodate more patients needing prostate MRI scans.
“Our study shows how DLR can effectively reduce diffusion weighted image noise and how DLR may be harnessed to reduce the time of a multiparametric prostate MRI exam without sacrificing image quality. These findings have the potential to improve patient experience and optimize resource utilization in clinical practice,” the group concluded.
The commercial algorithm used for the study was GE Healthcare’s AIR™ Recon DL.
Learn more here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She joined Innovate Healthcare in 2021 and has since put her unique expertise to use in her editorial role with Health Imaging.