Multiparametric MR imaging is an important diagnostic and staging tool for prostate cancer. In general, advances in MR oncologic imaging have enabled earlier diagnosis and more personalized patient treatments, including surveillance and less invasive treatment methods.1
Our hospital specializes in interventional radiology and prostate surgery with more than 1,000 patients receiving MR/ ultrasound fusion prostate biopsy or robotic surgery in the last few years. Currently, prostate MR exams are performed on a SIGNA™ Premier 3.0T system, featuring capabilities such as the AIR™ Anterior Array (AA) Coil. Initially, our software version was SIGNA™Works AIR™ Edition, which includes MUSE, FOCUS DWI, MAGiC DWI and DISCO. Recently, we upgraded our software to SIGNA™Works AIR™ IQ Edition, featuring AIR™ Recon DL, a deep-learning-based reconstruction algorithm that improves SNR and image sharpness, and enables shorter scan times, and DISCO Star, a multi-phase, dynamic sequence that is motion robust and free-breathing for better patient compliance. Now, these combined capabilities provided a wing-to-wing solution for prostate imaging that improves image quality, the reporting process and patient comfort.
DISCO Star is also a 3D radial (stack- of-stars) sequence that acquires one continuous dynamic arterial phase for consistent, worry-free timing of the contrast arrival, which will vary from patient to patient depending on age, pathology and condition.
Clinical impact
In T2 and diffusion MR exams with AIR™ Recon DL, the scan times were shortened by nearly 50%, yet we achieved the same high-resolution settings. As important, SNR increased significantly. With the shorter exam protocols, the rate of motion artifact and recurrence decreased significantly.
As a result of implementing AIR™ Recon DL, high resolution bi-parametric prostate exams are completed in 15 minutes and multi-parametric exams are completed in 25 minutes, which is much shorter than the conventional MR protocols used previously. Our daily number of patients and patient satisfaction have increased, and even claustrophobic patients are experiencing fewer concerns, all as a result of the shorter exam times.
With DISCO Star, the physiological motion of the bladder and bowel artifacts were reduced significantly. Temporal resolution of the perfusion series is set for 5 seconds per phase, and we are scanning 36 phases. This has increased our overall pharmacokinetic assessments with GenIQ software as well.
The benefit of using a high-density coil coverage, specifically to achieve a small- FOV and high resolution for prostate imaging, was also examined. Demanding prostate anatomy and high- resolution imaging needs have previously been a challenge to maintain SNR. With the AIR™ Multi-Purpose Coil, we have observed a 10%-20% signal increase advantage in comparison to the AIR™ AA coil.
Research efforts
In our initial use of the MAGiC sequence to generate multiple contrasts from one scan, we found it was more effective than described in the literature in brain MR imaging with a 48-channel head coil and a high-gradient SIGNA™ Premier 3.0T MR system. Therefore, we tested whether we could achieve the same performance in prostate imaging with a similar approach, MAGiC DWI, that allows us to synthetically generate high b-value diffusion images. The results were impressive with the ability to obtain artifact-free, 1 x 1 x 3 mm spatial resolution images.
Next in our department is a study of T1 and T2 maps to determine if it would be useful for differentiating tumor and tumor-like lesions, or if it would be easier to distinguish the lesions with artificial intelligence (AI). Currently, there are not many national and international studies on this subject. We will investigate the effectiveness of the MAGiC sequence used with AI‡ by comparing digital pathology results and digital radiology findings in patients who underwent radical prostatectomy. Our research project has received a national grant as part of TUBITAK 1001 program.
At Acıbadem Maslak Hospital, we aim to develop and nurture academic collaboration between Acıbadem Mehmet Ali Aydınlar University Radiology, Urology and Pathology sciences, Acıbadem Medical Engineering, and Istanbul Technical University Electronic Engineering through this project. With the support of high-end technology from GE Healthcare and partners, we expect to deliver meaningful contributions to MR with AI in prostate imaging.
References
- Stabile A, Giganti F, Rosenkrantz AB, et al. Multiparametric MRI for prostate cancer diagnosis: current status and future directions. Nat Rev Urol. 2020 Jan;17(1):41-61.