American College of Radiology Initiatives on Prostate Magnetic Resonance Imaging Quality

Abstract

Magnetic resonance imaging (MRI) has become integral to diagnosing and managing patients with suspected or confirmed prostate cancer. However, the benefits of utilizing MRI can be hindered by quality issues during imaging acquisition, interpretation, and reporting. As the utilization of prostate MRI continues to increase in clinical practice, the variability in MRI quality and how it can negatively impact patient care have become apparent. The American College of Radiology (ACR) has recognized this challenge and developed several initiatives to address the issue of inconsistent MRI quality and ensure that imaging centers deliver high-quality patient care. These initiatives include the Prostate Imaging Reporting and Data System (PI-RADS), developed in collaboration with an international panel of experts and members of the European Society of Urogenital Radiology (ESUR), the Prostate MR Image Quality Improvement Collaborative, which is part of the ACR Learning Network, the ACR Prostate Cancer MRI Center Designation, and the ACR Appropriateness Criteria. In this article, we will discuss the importance of these initiatives in establishing quality assurance and quality control programs for prostate MRI and how they can improve patient outcomes.

Introduction

Data from several multicenter randomized clinical trials support the inclusion of multiparametric magnetic resonance imaging (mpMRI) in the guidelines for diagnostic workup of patients with suspected or confirmed prostate cancer.^1–3 These trials have shown that, when used as a pre-biopsy triage method in patients with an elevated serum prostate-specific antigen (PSA), mpMRI increases the detection of clinically significant prostate cancer (usually defined as Gleason Grade Group 2 or higher) and reduces the detection of insignificant cancers compared to conventional ultrasound-guided systematic prostate biopsy, while also decreasing the number of unnecessary biopsies.^1,2 Outside of clinical trials, however, the benefit of mpMRI can be compromised by quality gaps in the management pathway linked to the performance of MRI and MRI-guided interventions.⁴ These quality gaps include artifacts from motion and geometric distortion by rectal air that degrade image quality: i) low signal-to-noise ratio (SNR) due to the patient’s intrinsic factors (e.g., large body habitus) or scanner-related factors; ii) poor compliance with Prostate Imaging Reporting and Data System (PI-RADS) technical standards; iii) variability in image interpretation, often affected by confounding conditions such as benign prostatic hyperplasia and prostatitis; iv) biopsy sampling errors, which are mainly influenced by technique and operator expertise; iv) variability in the histopathology assessment of biopsy samples.

The American College of Radiology (ACR) has recognized these challenges and, over the years, established many initiatives in the field of prostate MRI aimed at improving the quality of care delivered to patients undergoing prostate cancer workup. This article lists these initiatives and describes their importance in establishing quality assurance and quality control programs.

PI-RADS Guidelines Attempts to reduce variability in prostate MRI performance date back to 2009, when a panel of prostate cancer experts from European centers conducted a consensus meeting to make recommendations on a standardized method for the performance, interpretation, and reporting of mpMRI.⁵ A few years later, in 2012, the ACR Prostate Imaging Reporting and Data System (PI-RADS) committee was formed by members of the ACR, an international working group for prostate mpMRI convened by the AdMeTech foundation, and the European Society of Urogenital Radiology (ESUR). The successful collaboration between these groups resulted in the creation of the PI-RADS guidelines. The landmark paper describing the second version of the system (PI-RADS v2), published in European Urology and awarded the journal’s best scientific paper of 2015, stated the specific aims of the guidelines: i) to establish minimum acceptable technical parameters for prostate mpMRI acquisition; ii) to simplify and standardize the terminology and content of radiology reports; iii) to facilitate the use of MRI data for targeted biopsy; iv) to develop assessment categories that summarize levels of suspicion or risk and that can be used to select treatment naïve patients for biopsies and management; v) to enable data collection and outcome monitoring; vi) to educate radiologists on prostate MRI reporting and reduce variability in imaging interpretations; vii) to enhance interdisciplinary communications with referring clinicians.⁶ That article helped disseminate the minimum technical standards for mpMRI acquisition reporting, providing a 5-point scoring system to convey the likelihood of clinically significant cancer presence based on mpMRI findings. Importantly, like other ACR RADS, PI-RADS provides radiologists and referring physicians a common language, increasing confidence in the accuracy and reliability of prostate MRI exams.⁷ PI-RADS is now an international standard incorporated into multiple healthcare guidelines, including the European and US urological society guidelines.^3,8–10 Beyond assisting physicians in triaging patients for prostate biopsy, a growing body of evidence highlights the usefulness of the prostate cancer phenotypes described in PI-RADS as a biomarker for aggressive forms of prostate cancer that require treatment due to an increased risk of progression and development of metastatic disease, recurrence after treatment, and cancer-related death (Fig. 1).¹¹

Fig 1.

66-year-old man with elevated PSA (10 ng/mL) on active surveillance for a Grade Group 1 prostate cancer found on a prior systematic biopsy. A-D: PI-RADS-compliant multiparametric prostate MRI exam with adequate resolution and SNR, with no artifacts affecting the interpretation of the images (PI-QUAL score 5). Axial T2-W image show a 1.8 cm focal lesion with low signal intensity on the left posterolateral peripheral zone at the mid gland and apex with a broad capsular contact (arrow, a). The lesion demonstrates markedly hyperintense signal on high b-value (1,400 sec/mm²) DWI (arrow, b), markedly hypointense signal on ADC map (arrow, c), and early arterial enhancement on T1-weighted DCE images (arrow, d)*. The PI-RADS assessment category of the lesion is 5. MRI-guided biopsy of the lesion revealed prostate cancer Grade Group 2. This case illustrates how high-quality prostate MRI and MRI-guided biopsy can be helpful for the detection of clinically significant prostate cancer.

* PI-RADS guidelines recommend T1-weighted DCE images with fat suppression and/or subtractions. The subtraction images performed in this case are not shown.

PSA: prostate-specific antigen; PI-RADS: prostate image reporting and data system; SNR: signal-to-noise ratio; T2-W: T2-weighted; DWI: diffusion-weighted imaging; ADC: apparent diffusion coefficient; DCE: dynamic contrast-enhanced. PI-QUAL: prostate imaging quality.

The current version of the guideline (PI-RADS v2.1), released in 2019, addresses some inconsistencies and limitations of PI-RADS v2.¹² It includes clarifications and revisions of image data acquisition and interpretation criteria, a statement about biparametric MRI using T2-weighted and diffusion-weighted images, and structured report templates.¹³

In 2021, the PI-RADS steering committee released a consensus statement describing how the MRI-directed pathway for patients with a clinical suspicion of harboring prostate cancer could be implemented into clinical practice. ¹⁴ In that article, the steering committee also emphasized that, for MRI and MRI-directed biopsy to deliver the intended benefits, the quality of the entire diagnostic process must be ensured by robustly trained technologists, experienced radiologists, and practitioners who conduct MRI-directed biopsy while working within multidisciplinary teams.

Prostate MRI Education

Prostate MRI interpretation can be challenging and requires education, training, and awareness of common pitfalls, such as prostate cancer mimickers and imaging artifacts. Prostate MRI is now a component of training in most US-based abdominal and body MRI fellowships. Studies investigating the learning curve in prostate MRI interpretation have shown that a traditional didactic curriculum and exposure to a sufficient volume of exams can improve the readers’ performance.¹⁵ Formal instruction through didactic lectures, interactive case-based reviews, participation in multidisciplinary tumor boards, and interpretation feedback can improve prostate MRI interpretation. Rosenkrantz et al. showed that novice readers rapidly improve prostate cancer detection accuracy after initial exposure to prostate MRI exams, with the improvement starting to slow down after approximately 40 exams.¹⁶ Readers who received feedback after interpretation had significantly larger improvements in sensitivity for the detection of transitional zone tumors and a greater increase in confidence in interpreting compared to readers who did not receive feedback. Beyond clinically significant cancer detection, continued hands-on experience with case-based discussions and expert feedback has also been found to help readers improve their accuracy in prostate cancer staging.¹⁶

Applying PI-RADS assessment criteria and certain aspects of imaging acquisition has challenges. For this reason, the ACR provides educational material describing the major revisions in technical parameters for image acquisition and clarifications in the interpretation criteria described in PI-RADS v2.1.¹⁷ To supplement that material, a primer for using PI-RADS v2.1 was released in 2021. This self-paced, educational online module is available to all at no charge on the ACR website.¹⁸ It includes six sections covering prostate MRI anatomy, PI-RADS assessment principles, caveats for prostate MRI interpretation, prostate cancer staging with MRI, prostate MRI reporting, and basic technical parameters (Fig. 2).

Figure 2.

Primer for using PI-RADS v2.1 for Prostate MRI. The primer contains six modules that can be done nonlinearly and a page with links to additional educational resources. At the end of each module, a short quiz highlights important teaching points. The module is freely available on the ACR website, and no membership is required.²⁰

The ACR Education Center provides two offerings for healthcare professionals interested in receiving hands-on experience in prostate MRI interpretation. An introductory virtual micro-course provides participants access to recorded lectures and a collection of 20 prostate MRI exams curated with radiology-pathology feedback.¹⁹ A more immersive experience is offered at the ACR Education Center in Reston, Virginia, three times a year.²⁰ In this in-person, 2-day workshop, a broad range of topics is covered by didactic lectures and illustrative cases that participants can review at their individual workstations with assistance from supervising experienced faculty, many of whom are current or former members of the ACR PI-RADS steering committee (Fig. 3). Participants have access to over 150 prostate MRI exams from various US institutions, using different MRI equipment and imaging acquisition protocols, exposing participants to a broad spectrum of pathologies, including common and uncommon prostate conditions, pitfalls for imaging interpretation, and artifacts. Various techniques for MRI-guided biopsy and the basic steps for preparing the images for the procedure are reviewed in the course. Participants are required to complete at least 100 exams to receive an “ACR Certificate of Proficiency”. Since its introduction in 2016, over 500 radiologists from multiple countries have attended and successfully completed the prostate MRI workshop at the ACR Education Center.

Figure 3.

A. Facade of the ACR Education Center in Reston, VA. B. In the classroom, participants have individual workstations equipped with diagnostic grade monitors and access to the ACR case engine containing over 150 full prostate MRI datasets with radiology-pathology correlation. The ACR designates this live activity for a maximum of 20.5 American Medical Association Physician Recognition Award (AMA PRA) Category 1 Credits.

ACR Prostate MR Image Quality Improvement Collaborative

Achieving adequate image quality is the first step in the quality chain for prostate MRI.⁴ Even though PI-RADS guidelines helped disseminate the minimum technical standards for imaging acquisition, there is still considerable variability in adherence to those standards. This adherence is lower in non-academic centers and institutions that lack prostate MRI expertise.²¹ Furthermore, signal-to-noise ratio limitations, and artifacts from motion and geometric distortion of diffusion-weighted imaging/apparent diffusion coefficient maps by susceptibility artifacts can compromise the image quality even when PI-RADS standards are followed (Fig. 4).²²

Figure 4.

A3 Template used by the ACR Learning Network Prostate MR Image Quality Improvement Collaborative. Adapted from Larson DB, Mickelsen LJ, Garcia K. Realizing Improvement through Team Empowerment (RITE): A Team-based, Project-based Multidisciplinary Improvement Program. Radiographics. 2016 Nov-Dec;36(7):2170–2183.

Recognizing that high-quality prostate MRI can be consistently obtained in the setting of a well-managed quality assurance program, the ACR leadership created the Prostate MR Image Quality Improvement Collaborative as part of the ACR Learning Network. The Collaborative aimed to develop a standardized system to increase the number of MRI exams that meet quality criteria according to a modified version of the Prostate Imaging Quality (PI-QUAL) scale, which was established to grade the diagnostic adequacy of prostate MRI exams to rule in and rule out all clinically significant prostate cancer lesions on MRI.²³ In this program, participating organizations take advantage of the ACR ImPower structured program for quality improvement and work together to conduct an in-depth analysis of the current state of prostate MR image quality at their sites while learning how to use the A3 thinking method for problem-solving²⁴ (Fig. 5). This method allows teams to identify barriers to achieving high-quality prostate MR images and to design and implement sustainable solutions through “plan-do-study-act” cycles while addressing key drivers of prostate MR image quality with active participation from their frontline staff.^25,26 To improve de consensus on prostate MRI quality, a virtual training session and a review article on how to use PI-QUAL were provided to both technologists and radiologists involved in the evaluation of the quality of prostate MRI exams.²⁷

Figure 5.

A 71-year-old man with an elevated PSA of 7.29 ng/mL and no prior prostate biopsy. A. Axial T2-W image obtained in compliance with PI-RADS technical standards shows poor SNR and blurring artifact from rectal peristalsis resulting in a poor delineation of the prostate capsule and of the zonal anatomy. B-C: Axial high b value (1,500 sec/mm²) DWI and ADC map show susceptibility artifacts from rectal gas, distorting the signal on the prostate. D: T1-weighted DCE image obtained with a temporal resolution of 30 sec, which exceeds the 15-sec upper threshold recommended in PI-RADS. It was possible to identify a focal lesion with markedly restricted diffusion in the anterior transition zone, but it was not possible to rule in and rule out focal lesions in other regions of the prostate on DWI. Furthermore, adequate characterization of the focal lesion was not possible on T2-W images and DCE. Because none of the sequences had sufficient diagnostic quality, the overall PI-QUAL score of this exam is 1, and a PI-RADS score should not be assigned for the focal lesion identified in this setting.

PSA: prostate-specific antigen; T2-W: T2-weighted; PI-RADS: prostate image reporting and data system; SNR: signal-to-noise ratio; DWI: diffusion-weighted imaging; ADC: apparent diffusion coefficient; DCE: dynamic contrast-enhanced. PI-QUAL: prostate imaging quality.

In the program’s first cohort that concluded in December of 2022, all organizations identified opportunities to improve quality by reducing process variations across their sites. Those that implemented the interventions achieved their improvement goals, with substantial improvement in aggregated measured image quality for the sites as a whole.

The organizations that complete the program will have a chance to join the Learning Network and continue to exchange knowledge with subsequent cohorts to spread best practices related to improving prostate MR image quality broadly.

ACR Prostate Cancer MRI Center Designation

Concerns about variability in prostate MRI quality have been a significant barrier to its adoption in the diagnostic pathway for prostate cancer. Besides inconsistencies in image quality and poor compliance with PI-RADS technical standards, variations in readers’ performance in interpreting the images can negatively affect patient care. To address these concerns, professional organizations in Europe and in the UK, have established several initiatives to mitigate variability in the quality chain of prostate MRI.²⁸ In 2018, the ACR created a workgroup to set standards for a Prostate Cancer MRI Center Designation.²⁹ The designations’ goal is to recognize centers committed to offering high-quality prostate MRI beyond simply following the minimum technical standards described in PI-RADS.

Using the framework previously employed for the ACR breast cancer and lung cancer screening centers of excellence, the workgroup defined requirements for sites to receive the designation, contingent on ACR body MRI accreditation (Supplemental material). The initial qualification requires that radiologists interpret at least 150 exams independently, or 100 exams in a supervised setting, over the prior 36 months, to demonstrate sufficient experience with prostate MRI. The requirement for renewal is to interpret at least 100 exams in the prior 36 months. The designation requires centers to be able to perform MRI-guided biopsy of any kind (cognitive, MRI/ultrasound fusion biopsy, or in-bore biopsy) or to have an agreement with a referring center that can perform the biopsy. Centers must also have a mechanism for follow-up biopsy results. This mechanism can identify cases with radiology-pathology discordance, to help radiologists refine their interpretation skills and to allow facilities to establish provider and practice-level performance benchmarks for prostate cancer detection with MRI (Fig. 6).

Figure 6.

A 77-year-old man with an elevated PSA of 6.95 ng/mL, gland volume 99 cc, and no prior prostate biopsy, underwent a biparametric prostate MRI. A. Axial T2-W image obtained with a PI-RADS-compliant technical parameters shows blurring of the prostate due to motion artifacts. A focal lesion with hypointense signal is noted on the left posterior peripheral zone (arrow). B-C. Axial calculated high b-value (b 1,400 sec/mm²) DWI and ADC map have low SNR and susceptibility artifacts from rectal gas, resulting in distortion of peripheral zone signal. D-E. The technologist recognized the limitations of DWI/ADC images and obtained the DWI and ADC map using the RESOLVE technique, which improved the SNR and decreased the geometric distortion of the prostate, making it more apparent the presence of a focal lesion with markedly hyperintense signal on calculated high b-value (1,400 sec/mm²) DWI (arrow, d) and markedly hypointense signal on ADC map (arrow, d) in the left apex/mid posterior peripheral zone. Since DWI/ADC and T2W images taken together were considered to have sufficient diagnostic quality, the exam was rated as PI-QUAL 3.

The original interpretation of the images by a novice reader with 9 months of post-fellowship experience was PI-RADS score 2. Before the biopsy, the images were reviewed by a radiologist with 10 years of post-fellowship training who detected the left peripheral zone lesion and scored as PI-RADS 5 based on the DWI/ADC appearance. An MRI-guided biopsy of the lesion revealed Grade Group 3 prostate cancer with intraductal carcinoma features. Biopsy results were shared with the novice reader. This case highlights the role experienced technologists and radiologists play in quality assurance for prostate MRI and how poor image quality may impact inter-reader variability.

PSA: prostate-specific antigen; T2-W: T2-weighted; DWI: diffusion-weighted imaging; ADC: apparent diffusion coefficient; SNR: signal-to-noise ratio; Resolve: readout segmentation of long variable echo-trains. PI-QUAL: prostate Imaging quality.

Facilities applying for the designation must submit examples of prostate MRI exams to evaluate exam quality, including the absence of artifacts and compliance with PI-RADS technical standards. The only technical parameter the designation differs from PI-RADS v2.1 standards is the spatial resolution requirement (Table 1). Instead of setting maximum in-plane pixel dimensions in the phase and frequency directions, the designation requires a maximum pixel area with less stringent criteria deemed necessary to maintain an adequate signal-to-noise ratio in some cases.³⁰ For instance, PI-RADS v.2.1 requires T2-weigthed images with in-plane dimensions ≤ 0.7 mm and ≤ 0.4 mm in the phase and frequency directions, respectively, while the designation sets a maximum pixel area of 0.6 mm².

Table 1.

ACR Prostate Cancer MRI Center Designation technical parameters criteria for prostate mpMRI. FOV: field of view; TE: echo time; TR: Repetition time.

MR prostate exam
Required Sequences	Category A: Pulse Sequence and Image Contrast	Category B: Anatomic Coverage and Imaging Planes	Category C: Spatial Resolution
Multiplanar T2W	•  2D T2W Fast-spin-echo (FSE) or Turbo-spin-echo (TSE) •  3D T2W can be used as an adjunct to 2D	•  FOV: 12–20 cm to encompass the entire prostate and seminal vesicles •  Should include axial plane (either straight axial to the patient or in an oblique axial plane matching the long axis of the prostate) and a minimum of one additional orthogonal plane (i.e., sagittal and/or coronal)	•  Slice thickness: ≤ 3 mm •  Gap: 0 mm •  Pixel area: ≤ 0.6 mm2
Axial diffusion weighted images (DWI) with apparent diffusion coefficient (ADC) maps	•  Free-breathing spin echo-echo planar imaging (EPI) combined with spectral fat saturation is recommended •  TE ≤ 90 msec •  TR ≥ 3000 msec •  b-values: low, intermediate and high b-values required for ADC map creation (low b-value 0–100 sec/mm2 and intermediate value 800–1000 sec/mm2); additional high b-values image (≥1400 sec/mm2) should be acquired separately or calculated from images obtained with low and intermediate b-values < 1000 sec/mm2 Note: low, intermediate and high b-value images must be submitted as separate sequences.	•  FOV 16–22 cm •  Axial (same as T2W)	•  Slice thickness: ≤ 4 mm •  Gap: 0 mm •  Pixel area: ≤ 10 mm2
Axial pre-contrast and dynamic contrast enhanced (DCE)	•  Must have pre-contrast images •  T1W gradient echo (GRE) •  3D is preferred over 2D •  Fat suppression and/or subtractions recommended •  TE <5 msec; TR <100 msec •  Contrast dose: 0.1 mmol/kg standard Gadolinium-based contrast agent •  Injection rate: 2–3 mL/sec starting with continuous image data acquisition (should be the same for all exams)	•  FOV: should encompass the entire prostate gland and seminal vesicles •  Axial (same as T2W and DWI)	•  Slice thickness: ≤ 3 mm •  Gap: 0 mm •  Temporal resolution: ≤15 sec •  Acquisition time: ≥ 2 min •  Pixel area: ≤ 4.0 mm2

ACR Appropriateness Criteria

During the 1990s, the ACR recognized the need to define national guidelines for the appropriate use of imaging technologies, known as the ACR Appropriateness Criteria (ACR AC)³¹. The ACR AC is an evidence-based tool for assessing the appropriateness of imaging procedures for specific clinical scenarios and was designed to promote quality, safety, and efficacy in imaging. Recommendations made by the ACR AC are based on a rigorous review of the medical literature and are supplemented by expert opinion and clinical experience.

In June 2016, the Centers for Medicare & Medicaid Services designated the ACR a “qualified Provider-Led Entity” approved to provide appropriate use criteria under the Medicare Appropriate Use Criteria program for advanced diagnostic imaging. That designation allowed medical practitioners to consult the ACR AC to fulfill requirements specified in Protecting Access to Medicare Act before ordering advanced diagnostic imaging for Medicare patients.³¹

The ACR AC is periodically revised to address new literature evidence that becomes available. The appropriateness of prostate MRI for prostate cancer has evolved in the revisions of the ACR AC. In the most recent version released in 2022, the topic “Pre-treatment detection, surveillance and staging of prostate cancer” covers five clinical scenarios.³² MpMRI (“MRI pelvis without and with intravenous contrast”) and biparametric MRI (“MRI without intravenous contrast”) of the pelvis are designated as “usually adequate” for patients with clinically suspected prostate cancer without or with prior negative prostate biopsy, and for patients with clinically established low-risk prostate cancer on active surveillance. MpMRI is also “usually appropriate,” whereas biparametric MRI is designated “may be appropriate” for patients with clinically established intermediate or high-risk prostate cancer in the context of staging.

On the topic “Post-Treatment Follow-up of Prostate Cancer,” the panel recognized the benefits of dynamic contrast-enhanced imaging for detecting residual or recurrent prostate cancer in patients after radical prostatectomy and nonsurgical local and pelvic treatments (e.g., radiation and ablation therapies).³² In this clinical scenario, mpMRI of the pelvis is “usually appropriate,” whereas biparametric MRI is designated as “may be appropriate.” The panel also stated that pelvic MRI is complementary to positron emission tomography (PET) examinations using PSMA-radioligands, choline, and Fluciclovine in this setting, and both categories of examinations may be beneficial to perform. For the follow-up of patients with metastatic prostate cancer treated by systemic therapy (androgen deprivation therapy, chemotherapy, immunotherapy), the likelihood of metastatic disease outside the pelvis is higher; therefore, evaluation of local recurrence becomes of lesser clinical importance unless it is locally advanced and is causing urinary or bowel complications. In this setting, mpMRI of the pelvis is considered “may be appropriate”, while biparametric MRI is considered “usually not appropriate”.

Conclusions

The ACR has developed impactful initiatives encompassing all aspects of prostate MRI quality, from its appropriate use to adequate performance and accurate interpretation, to improve the quality of care delivered to patients with suspected or confirmed prostate cancer. These initiatives will continue to evolve as new evidence becomes available in this field. While some initiatives only apply to facilities in the United States, such as the Prostate Cancer MRI Center Designation, their framework is publicly available and can be used freely across the world.