• AI
  • Molecular Imaging
  • CT
  • X-Ray
  • Ultrasound
  • MRI
  • Facility Management
  • Mammography

Can Deep Learning Reduce DWI MRI Exam Time Without Affecting Prostate Image Quality?

News
Article

Employing deep learning reconstruction at four excitations for DWI MRI may lead to an average five-minute reduction in exam time for prostate mpMRI, according to a new study.

For prostate diffusion-weighted imaging (DWI) magnetic resonance imaging (MRI), researchers suggest that obtaining deep learning reconstruction (DLR) images at reduced excitations may lead to a greater than 65 percent reduction in scanning time for restricted field of view (FOV) image acquisition.

For the retrospective study, recently published in Clinical Imaging, researchers examined the impact of DLR (AIR™ Recon DL, GE HealthCare) at a lower number of excitations (NEX) upon image quality, signal-to-noise ratio (SNR) and examination time for DWI MRI in 52 patients (median age of 64 and median prostate-specific antigen (PSA) level of 5.9 ng/mL).

The study authors found that utilizing DLR at 4NEX for DWI MRI reduced scanning time by 68 percent for restricted FOV images and 39 percent for large FOV image series. Overall, the researchers estimated that obtaining 4NEX DLR images for DWI MRI would take 13 minutes and 37 seconds in comparison to 18 minutes and 37 seconds for standard imaging acquisition with prostate mpMRI.

Can Deep Learning Reduce DWI MRI Exam Time Without Affecting Prostate Image Quality?

Here one can see restricted and large field-of-view (FOV) MRI images obtained at variable excitations for a patient with a lesion in the right mid-gland peripheral zone. New research suggests that utilizing deep learning reconstruction at 4NEX for DWI MRI may reduce scanning time by 68 percent for restricted FOV images and 39 percent for large FOV image series. (Images courtesy of Clinical Imaging.)

“This time savings has the potential to translate to ≥1 additional patients imaged per day per scanner, assuming a routine time slot of 20–30 min and 4–6 mpMRI prostate examinations,” wrote lead study author Rory L. Cochran, M.D., Ph.D., who is affiliated with the Department of Radiology at Massachusetts General Hospital in Boston, and colleagues.

In comparison to the SNR for conventional NEX with restricted FOV DWI MRI images (7.8), the researchers noted higher SNR with DLR at 6NEX (9.1) and 9NEX (10.1) but no significant difference at 4NEX (8.2).

Three Key Takeaways

1. Significant reduction in MRI scanning time. Using deep learning reconstruction (DLR) for prostate DWI MRI with reduced excitations can cut scanning time by over 65 percent for restricted field of view (FOV) imaging, potentially allowing for more patients to be scanned each day.

2. Comparable image quality at lower excitations. The study found no significant difference in image quality between DLR at 4 excitations (NEX) and conventional methods, suggesting that reduced excitations may maintain diagnostic accuracy while improving efficiency.

3. Signal-to-noise ratio (SNR) insights. DLR imaging at higher excitations (6NEX and 9NEX) achieved a higher SNR compared to conventional methods for restricted FOV images with the exception of a minimal difference at 4NEX. This balance of SNR and efficiency supports DLR use, particularly in restricted FOV imaging.

(Editor’s note: For additional content on prostate cancer imaging, click here.)

For large FOV DWI MRI, regardless of NEX level, the study authors noted a 19.1 SNR for conventional NEX vs. a 19.6 SNR for 4NEX DLR at 4NEX and a 21.3 SNR for 6NEX DLR.

“No significant difference was detected between conventional method reconstructed images and DLR diffusion images using 4 NEX datasets … . SNR for the large FOV DWI images was not significantly different for DLR images compared to conventional reconstruction … , regardless of NEX,” added Cochran and colleagues.

(Editor’s note: For related content, see “AI Segmentation, Intraprostatic Tumor Volume and Metastases: What a new mpMRI Study Reveals,” “Emerging AI Platform Shows Promise for Prostate Cancer Detection on mpMRI” and “MRI Study Suggests Deep Learning Model Offers Equivalent Detection of csPCa as Experienced Radiologists.”)

Beyond the inherent limitations of a retrospective study, the authors cautioned that utilizing reduced NEX may increase the number of non-diagnostic diffusion images.

Recent Videos
Can 18F-Floutufolastat Bolster Detection of PCa Recurrence in Patients with Low PSA Levels After Radical Prostatectomy?
Detecting PCa Recurrence in African Americans: Can 18F-Flotufolastat Have an Impact?
Emerging Research at SNMMI Examines 18F-flotufolastat in Managing Primary and Recurrent Prostate Cancer
Could Pluvicto Have a Role in Taxane-Naïve mCRPC?: An Interview with Oliver Sartor, MD
A Closer Look at MRI-Guided Transurethral Ultrasound Ablation for Intermediate Risk Prostate Cancer
Nina Kottler, MD, MS
Brian F. Chapin, MD
Related Content
© 2024 MJH Life Sciences

All rights reserved.