Abstract
OBJECTIVE.
The purpose of this study was to report on the practice patterns and challenges of performing and interpreting prostate MRI.
SUBJECTS AND METHODS.
An electronic survey regarding prostate MRI practice patterns and challenges was sent to members of the Society of Abdominal Radiology.
RESULTS.
The response rate was 15% (212/1446). Most (65%) of the respondents were academic abdominal radiologists with 1–5 (52%), 6–10 (20%), 11–20 (15%), and more than 20 (5%) years of experience in reporting prostate MRI. The numbers of prostate MRI examinations reported per week were 0–5 (43%), 6–10 (38%), 11–20 (12%), 21–30 (5%), and more than 30 (2%). Imaging was performed at 3 T (58%), 1.5 T (20%), or either (21%), and most examinations (83%) were performed without an endorectal coil. Highest b values ranged from 800 to 5000 s/mm2; 1400 s/mm2 (26%) and 1500 s/mm2 (30%) were the most common. Most respondents (79%) acquired dynamic contrast-enhanced images with temporal resolution of less than 10 seconds. Most (71%) of the prostate MRI studies were used for fusion biopsy. PI-RADS version 2 was used by 92% of the respondents and template reporting by 80%. Challenges to performing and interpreting prostate MRI were scored on a 1–5 Likert scale (1, easy; 2, somewhat easy; 3, neutral; 4, somewhat difficult; 5, very difficult). The median scores were 2 or 3 for patient preparatory factors. Image acquisition and reporting factors were scored 1–2, except for performing spectroscopy or using an endorectal coil, both of which scored 4. Acquiring patient history scored 2 and quality factors scored 3.
CONCLUSION.
Most radiologists perform prostate MRI at 3 T without an endorectal coil and interpret the images using PI-RADS version 2. Challenges include obtaining quality images, acquiring feedback, and variability in the interpretation of PI-RADS scores.
Keywords: MRI, PI-RADS, practice patterns, prostate, survey
Prostate cancer is second most common malignancy in men worldwide [1]. Technical advances in prostate MRI have led to improvements in the diagnosis of prostate cancer with multiparametric imaging, which combines anatomic T2-weighted imaging with functional techniques, such as DWI and dynamic contrast enhancement (DCE) [2–5]. This has expanded the indications from local staging to include detection, localization, characterization, risk stratification, surveillance, assessment of recurrence, and image guidance for biopsy and treatment, such as prostatectomy and radiation and focal therapies [6–10].
The increased use and availability of prostate MRI have led to the development of PI-RADS, which was introduced in 2012. An updated version (PI-RADS v2) was published in 2016, and further refinements (PI-RADS v2.1) were made in 2019 [11–13]. The aim of PI-RADS is to standardize acquisition, interpretation, and reporting of prostate MRI to improve detection, localization, characterization, and risk stratification. PI-RADS is a working document. Although PI-RADS v2 has been validated [14], practice patterns continue to vary, and challenges in performing and interpreting prostate MRI persist [15, 16]. Furthermore, prostate MRI is being used in the PI-RADS MRI-directed biopsy pathway to increase detection of clinically significant prostate cancer, decrease detection of clinically insignificant cancer, and recognize patients who may not need a prostate biopsy [17].
The objective of this study was to report on the practice patterns and challenges of performing and interpreting prostate MRI examinations. A knowledge of the practice patterns and challenges will enable us to design education programs and guide further refinements for the next version of PI-RADS.
Subjects and Methods
This prospective survey-based study was conducted after approval was received from the research ethics office of the University of British Columbia. Completion of the survey served as consent to participate. The study was developed by the authors, who are members of the Society of Abdominal Radiology Prostate Disease–Focused Panel.
The survey consisted of 25 questions. Most of the questions were multiple choice with the option of selecting Other to include open-ended answers. One question had 16 parts with responses of Yes, No, or Sometimes. Another question on the challenges of performing and interpreting prostate MRI had 25 parts with Likert scales as responses. The respondents were asked to rate the difficulty of various aspects in the categories of patient preparation, technique, access to clinical information, and quality assurance. The respondents were asked to rank them as 1, easy; 2, somewhat easy; 3, neutral; 4, somewhat difficult; or 5, very difficult. The Schulze method was used to determine the least to most difficult items. A median score was calculated to provide additional information on whether an item was generally regarded as easy or difficult. Two open-ended questions at the end of the survey concerned challenges not mentioned in the survey and largest challenges. The survey was designed to be completed in 15 minutes.
The survey was administered electronically to all members of the Society of Abdominal Radiology on September 13, 2017. It was distributed on the FluidSurveys platform. The survey was anonymous, and there was no incentive to complete it. The survey was intended for radiologists who report or plan to report prostate MRI. The respondents were allowed to leave questions unanswered. Follow-up e-mail reminders to complete the survey were sent on October 3 and October 19, 2017. The survey reached 1446 members of the Society of Abdominal Radiology. It was closed November 30, 2017. The data were managed with Excel for Windows software (version 2016, Microsoft) and summarized with descriptive statistics.
Results
Survey Population
The response rate for this survey was 15% (212/1446). Characteristics of the survey population are shown in Table 1. Most of the respondents were practicing in the United States (159 [75%]) and Canada (37 [17%]). Most of the respondents were academic (65%) abdominal radiologists (85%) interpreting abdominal (97%) and prostate (91%) MRI. Although radiologists at different career stages were well represented, the largest group (52%) had been reading prostate MRI studies for only 1–5 years, followed by 6–10 years (43%), 11–20 years (31%), greater than 20 years (11%), and 18% had not yet started.
TABLE 1:
Survey Population (n = 212)
| Characteristic | % |
|---|---|
| Practice location | |
| United States | 75 (159) |
| Canada | 17 (37) |
| Central and South America | 3 (6) |
| Asia | 3 (6) |
| Europe | 2 (4) |
| Practice setting | |
| Academic | 65 (138) |
| Community | 19 (40) |
| Both academic and community | 13 (28) |
| Other (Veterans Affairs, nonprofit, private hospital) | 3 (6) |
| Type of practice | |
| Abdominal radiology | 85 (180) |
| Interpret abdominal MRI | 97 (206) |
| Interpret prostate MRI | 91 (193) |
| Years in practice | |
| 0–5 | 30 (63) |
| 6–10 | 21 (44) |
| 11–15 | 22 (46) |
| 11–20 | 9 (19) |
| > 20 | 19 (40) |
| Years of experience in reporting prostate MRI | |
| 0 (not yet) | 8 (18) |
| 1–5 | 52 (110) |
| 6–10 | 20 (43) |
| 11–20 | 15 (31) |
| > 20 | 5 (11) |
Note—Values in parentheses are numbers of respondents.
Preferences on Learning Prostate MRI
Eighteen (8%) respondents were not currently interpreting prostate MRI. Their intentions about whether or when to train and their preferred formats for training are shown in Table 2. Most preferred to train in a workshop format. Among respondents already interpreting prostate MRI (n = 194), however, the preferences for receiving continuing education updates were web based (60%), workshop (33%), and other (7%).
TABLE 2:
Preferences on Learning Prostate MRI
| Preference | Percentage |
|---|---|
| Plans for learning prostate MRI (n = 18) | |
| Do not plan to train | 17 (3) |
| Not sure when they plan to train | 22 (4) |
| Plan to train within 6 mo | 17 (3) |
| Plan to train in 6–12 mo | 22 (4) |
| Plan to train in 12–24 mo | 22 (4) |
| Format for training (n = 15) | |
| Workshop | 60 (9) |
| Web based | 33 (5) |
| Other | 7 (1) |
Note—Values in parentheses are numbers of respondents.
Practice Patterns in Prostate MRI
Practice patterns are shown in Table 3. The largest group for number of prostate MRI studies interpreted a week was 0–5 (43%), followed by 6–10 (38%). Twenty-two of these respondents were community radiologists, and all but one reported zero to five cases per week. The other community radiologist reported 21–30 cases per week. Eighteen community radiologists were currently not reporting prostate MRI, and all but three planned to train in this field. Most of the referrals for prostate MRI came from urologists within the institution (37%). Most of the respondents (64%) did not have waiting lists; 27% did. The waiting lists varied from 2–3 days to 1.5 years These results varied by country. Among radiologists practicing in Canada (37 responses), 68% reported having a waiting list; 74% of 139 radiologists practicing in the United States reported not having a waiting list.
TABLE 3:
Practice Patterns
| Characteristic | Percentage |
|---|---|
| No. of prostate MRI studies reported per week (n = 194)a | |
| 0–5 | 43 (84)a |
| 6–10 | 38 (73) |
| 11–20 | 12 (23) |
| 21–30 | 5 (10)a |
| 31–50 | 2 (3) |
| > 50 | 1 (1) |
| Referral pattern (n = 299)b | |
| Urologists within own institution | 37 (110) |
| Urologists outside of own institution | 30 (89) |
| Primary care | 9 (28) |
| Radiation oncology | 22 (67) |
| Other (general oncologist, internist, urooncologist, self referral, don’t know) | 2 (5) |
| Waiting list (n = 192) | |
| Yes | 27 (52) |
| No | 64 (123) |
| Sometimes | 8 (15) |
| MRI prostate examinations used for fusion biopsy (n = 201) | |
| Yes | 71 (143) |
| No | 21 (42) |
| Cognitive | 8 (16) |
| Who performs fusion biopsies (n = 158) | |
| Urologists | 74 (117) |
| Radiologists | 18 (28) |
| Both | 8 (13) |
Note—Values in parentheses are numbers of respondents.
Twenty-two were community radiologists; 21 reported zero to five cases per week, and one reported 21–30 cases per week.
More than one answer allowed.
Fifty-nine (28%) of the respondents reported using an endorectal coil and inflating it with air (61%), barium (23%), perfluorocarbon (10%), or another substance (6%), such as saline solution. The highest b value used for the DWI sequence ranged from 800 to 5000 s/mm2; 1500 s/mm2 was the most common (30%), followed by 1400 s/mm2 (26%) (Table 4).
TABLE 4:
Highest b value Used (n = 191)
| b (s/mm2) | Percentage |
|---|---|
| 800 | 4 (8) |
| 1000 | 9 (17) |
| 1200 | 5 (9) |
| 1250 | 0.5 (1) |
| 1400 | 26 (50) |
| 1500 | 30 (57) |
| 1600 | 11 (22) |
| 1700 | 0.5 (1) |
| 1800 | 0.5 (1) |
| 2000 | 13 (24) |
| 5000 | 0.5 (1) |
Note—Values in parentheses are numbers of respondents
In terms of adjusting the window and level for interpreting the ADC map, 71% of 198 respondents stated that they adjust the window and level subjectively with the mouse until they are satisfied; 18%, that they use a set window and level; 10%, that they use the PACS defaults; and 1%, that they use other methods, such as a commercially available computer-aided diagnosis program, starting with the defaults and then rarely adjusting if it may help and using automatic window and level settings using a box drawn in the center of the gland. Most of the respondents (150/200 [75%]) did not use computer-aided diagnosis programs. Among the 22% who did, DynaCAD (Philips Healthcare) was the most popular; iCAD (iCAD), Vitrea (Vital Images), Merge CAD-Stream (IBM), MIM Symphony (MIM Software), OsiriX (Pixmeo), PROCAD (PROCAD Software), and an in-house program were also used.
For 71% of the 201 respondents, prostate MRI was used for fusion biopsy, and for 21% it was not. The other 8% used prostate MRI for cognitive fusion. Most of the fusion biopsies were performed by urologists (74%), followed by radiologists (18%), and both (8%). The product most commonly used was UroNav (Philips Healthcare). Others, less common, products included Artemis (Eigen), Biobot (Biobot Analytics), BioJet (DK), MIM Symphony (MIM Software), Sentinelle (Invivo), and products from Koelis, Hologic, Siemens Healthineers, and Toshiba.
Ninety-seven (49%) respondents reported maintaining quality or auditing reports; 56 (28%) stated that they did not; and 46 (23%) reported that they sometimes did. Of these 199 respondents, 133 (67%) provided comments. Seven of the respondents did not have a mechanism in place to maintain quality, and two respondents replied that the question was not applicable. Common approaches for other 124 (93%) respondents included having reports double or triple read, monthly or biweekly meetings or rounds, follow-up pathology reports, follow-up with urologists, and using databases.
Respondents were asked for 16 items related to patient preparation technique and quality assurance and to report whether they use them by responding with yes, no, or sometimes (Figs. 1 and 2). These questions were answered by 177–200 respondents. Items such as access to 3 T imaging (82%), having patients empty their bladder (59%) or evacuate the rectum (54%) before MRI, using temporal resolution of less than 10 seconds (91%) and a b value of at least 1400 s/mm2 (82%), using PI-RADS v2 (92%), template reporting (80%), and having a tumor board (67%) had been adopted by most respondents. Other items not used or performed by most included having patients evacuate air in the rectum if present (73%), having patients refrain from ejaculation 2–3 days before the examination (65%), using an antispasmodic agent (59%), using an endorectal coil (68%), and performing spectroscopy (91%). In terms of audit and quality as well as use of extrapolated b values, the results were mixed.
Fig. 1—
Bar graph shows practice patterns in regard to patient preparation factors for performing prostate MRI. Values are number of responses with percentage in parentheses.
Fig. 2—
Bar graph shows practice patterns for performing and interpreting prostate MRI. Values are number of responses with percentage in parentheses.
Challenges in Performing and Interpreting Prostate MRI
The challenges in the preparatory factors for prostate MRI are shown in Figure 3. Having the patient empty their bladder before the MRI examination was the easiest to achieve (median score, 2). The other patient preparation factors all scored a median of 3; however, evacuating air in the rectum was considered the most challenging within this grouping. With regard to performing prostate MRI (Fig. 4), using PI-RADS v2 was considered the least challenging scoring (median, 1), whereas using an endorectal coil and performing spectroscopy were considered the most challenging, both receiving median scores of 4. All items with regard to obtaining patient information (Fig. 5) scored a median of 2, but obtaining the PSA level was the easiest, and obtaining a history of taking medication to reduce benign prostatic hyperplasia was the most challenging. For quality assurance items (Fig. 6), urologist interest was the least challenging (median score, 1), and performing audits and maintaining quality were the most challenging of the items (median score, 3).
Fig. 3—
Stacked bar graph shows challenges in patient preparation for prostate MRI in increasing order of difficulty from top to bottom according to Schultz method. Values are percentages, which may not total 100 owing to rounding.
Fig. 4—
Stacked bar graph shows challenges of performing prostate MRI with regard to technique listed in order of increasing difficulty from top to bottom according to Schultz method. Values are percentages, which may not total 100 owing to rounding.
Fig. 5—
Stacked bar graph shows challenges of performing prostate MRI with regard to obtaining patient information listed in order of increasing difficulty from top to bottom according to Schultz method. Values are percentages, which may not total 100 owing to rounding. BPH = benign prostatic hyperplasia.
Fig. 6—
Stacked bar graph shows challenges of performing prostate MRI with regard to quality assurance listed in order of increasing difficulty from top to bottom according to Schultz method. Values are percentages, which may not total 100 owing to rounding.
Respondents were asked to list any additional challenges that were not noted in the survey. Concern with PI-RADS was a common theme in the responses. Issues included interobserver variability in assigning scores, low PPV of PI-RADS category 4 lesions, PI-RADS for biopsy-proven prostate cancer, PI-RADS for younger patients at low risk with prostatitis, and low-risk PI-RADS scores being restrictive. Another theme was education or training. Comments included access to workshops for continuing education sometimes competing with academic responsibilities, having sufficient cases for training, and education levels affecting the quality of interpretations. Communication between the PACS and DynaCAD and UroNav programs and between radiologists and urologists was also mentioned. Funding and access to MRI, feedback and follow-up, and image quality were also mentioned.
For the final question of what are the greatest challenges with regard to prostate MRI, the main themes were access to MRI, having the appropriate magnet field strength and image quality, waiting lists, PI-RADS interobserver variability, quality of reporting and feedback, and ensuring radiologists have interest and training along with continuing education.
Discussion
The response rate for this survey was 15%, which is similar to or better than that of other radiology surveys [18–20]. Most of the respondents were new to reporting prostate MRI, having less than 5 years of experience. Most of those currently not reporting MRI planned to train in it. This pattern aligns with the rapidly increasing demand for prostate MRI [21] and will likely continue owing to increases in utilization to include MRI before biopsy [17, 22].
There was a lack of uniformity in performing prostate MRI. A fair number of respondents selected Sometimes as a response. For example, 10% of the respondents sometimes use an endorectal coil. This variation may be due to patient factors rather than local practice. Protocols may vary depending on indications, such as MRI for detection as opposed to stereotactic biopsy as opposed to surgical planning.
Nearly all of the respondents used PI-RADS v2 (92%) in their interpretations, but fewer followed PI-RADS v2 technique: 82% used highest b values of 1400 s/mm2 or greater, and 91% used temporal resolution of less than 10 seconds. It is unclear why these numbers were not larger. Eighteen percent of respondents did not use b value of 1400 s/mm2 or greater, which can lead to erroneous interpretations [23, 24]. One reason may be that the MRI system may not have the capability of generating quality images at the recommended b value threshold; in that case, extrapolated b values should be used. However, only a minority of respondents routinely used extrapolated b values, even though doing so does not incur additional time in the MRI unit. Possible reasons are that the MRI systems are not capable of this b value and that radiologists or technologists do not know how to access this option or are unaware of it. MRI systems with low motion probing gradients take a longer time to generate high b values, resulting in long TE and low image quality.
Another reason for not using appropriate temporal resolution of less than 10 seconds, in addition to technical limitations, is that DCE is not the dominant sequence for interpreting lesions and can only be used to upgrade a lesion in the peripheral zone, if it is a PI-RADS category 3 lesion on DWI with positive DCE results. However, the compliance rate we found for adherence to DCE temporal resolution of less than 10 seconds was higher than that of prior studies [25, 26].
Variation in degree of compliance with technical imaging parameters in PI-RADS v2 has been reported [25–27]. The causes are not apparent and may be related to MRI system incapability [24]. Further causes of lack of adherence to technical specifications may be different protocols on different MRI systems and radiologist or MRI technologist preference. Our study adds to the data that stronger efforts to educate radiologists regarding adoption of minimal technical standards are required to produce higher-quality examination, which in turn should result in improved interpretation and outcomes. Implementation of these standards should also be considered for site accreditation.
The use of template reporting (80%) was low. A possible reason may be technical challenges or limitations in incorporating a template onto the varied and currently available dictation and voice recognition platforms. Structured reporting has been found to be a strong preference for referring urologists [20]. Thus, template reporting should be more strongly considered and emphasized in the next iteration of PI-RADS.
Most of the prostate MRI examinations were used for fusion biopsy (78%), and the fusion biopsies are performed mainly by urologists (74%). Similar results were found in a study that surveyed urologists on their use of targeted MRI- or ultrasound-guided biopsy [28]. That study showed that urologists support the use of fusion biopsy and that 72% of academic urologists use it in their practice.
One of the common challenges was the interobserver variability of PI-RADS scores, which is a known limitation [15]. The release of PI-RADS v2.1 aimed to address some of these issues; radiologists were encouraged to validate and further improve interobserver variability in subsequent versions [13]. Acquiring and maintaining skills were also challenges. Some respondents lacked a feedback mechanism, and some thought that they did not have sufficient cases to maintain their skills. Most of respondents reported interpreting fewer than 10 cases per week. Almost one-half (47%) of respondents interpreted zero to five cases per week, and 38% interpreted 6–10 cases per week. It has been found [25, 29] that more experienced readers have better performance and less interobserver variability. One solution is to have a dynamic web-based resource and repository. The onus of maintaining quality reporting in the ever-increasing volume of cases should drive even the busiest of radiologists to take advantage and use these resources.
Although there have been surveys on preferences and content in prostate MRI reports [20, 30] and adherence to technical acquisition parameters [25–27], to our knowledge there have been no surveys encompassing the practice patterns and challenges in performing and interpreting prostate MRI. In conjunction with the explosive demand for prostate MRI, this information is valuable because it will help in development of impactful educational programs. These teaching resources should be designed to ensure proper training with a conscious effort to maintain quality [21].
There were several limitations to this study. This survey had a low response rate. Most of the respondents were academic abdominal radiologists practicing in North America, which might have introduced bias; thus, some of the practice patterns and challenges may not apply to other settings. Even at academic centers, there is variability in performance [31, 32]. Challenges are perceived to be greater for community radiologists performing and interpreting prostate MRI because of the absence of sub-specialty training, lack of familiarity with prostate MRI protocol implementation, and insufficient time. Thus, further education efforts regarding prostate MRI should focus on this group. Another potential limitation is that the survey results were obtained during the era of PI-RADS v2. Prostate MRI is a rapidly expanding field, and PI-RADS has been updated to PI-RADS v2.1 with minor overall changes [13].
In conclusion, most academic abdominal radiologists reported performing prostate MRI at 3 T without an endorectal coil and interpreted the images using PI-RADS v2 with a template. Most of the prostate MRI examinations were used for fusion biopsy, performed mostly by urologists. The greatest challenges were image quality, acquiring feedback, maintaining skills, and variability in the interpretation of PI-RADS scores. This information is important to guide educational programs and aid in further refinements in the next version of PI-RADS.
Footnotes
B. Turkbey has cooperative research development agreements with Philips Healthcare and Nvidia, receives royalties from the U.S. Government, and has a patent for related intellectual property in the field of prostate computer-aided diagnosis (National Institutes of Health–owned). The remaining authors declare that they have no disclosures relevant to the subject matter of this article.
The opinions and assertions contained herein are the private views of the authors and are not to be construed as official or as representing the views of the National Cancer Institute.
Based on a presentation at the Radiological Society of North America 2018 annual meeting, Chicago, IL.
References
- 1.Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018; 68:394–424 [DOI] [PubMed] [Google Scholar]
- 2.Fütterer JJ, Heijmink SW, Scheenen TW, et al. Prostate cancer localization with dynamic contrast-enhanced MR imaging and proton MR spectroscopic imaging. Radiology 2006; 241:449–458 [DOI] [PubMed] [Google Scholar]
- 3.Tanimoto A, Nakashima J, Kohno H, Shinmoto H, Kuribayashi S. Prostate cancer screening: the clinical value of diffusion-weighted imaging and dynamic MR imaging in combination with T2-weighted imaging. J Magn Reson Imaging 2007; 25:146–152 [DOI] [PubMed] [Google Scholar]
- 4.Kitajima K, Kaji Y, Fukabori Y, Yoshida K, Suganuma N, Sugimura K. Prostate cancer detection with 3 T MRI: comparison of diffusion-weighted imaging and dynamic contrast-enhanced MRI in combination with T2-weighted imaging. J Magn Reson Imaging 2010; 31:625–631 [DOI] [PubMed] [Google Scholar]
- 5.Franiel T, Stephan C, Erbersdobler A, et al. Areas suspicious for prostate cancer: MR-guided biopsy in patients with at least one transrectal US-guided biopsy with a negative finding—multiparametric MR imaging for detection and biopsy planning. Radiology 2011; 259:162–172 [DOI] [PubMed] [Google Scholar]
- 6.Hövels AM, Heesakkers RA, Adang EM, et al. The diagnostic accuracy of CT and MRI in the staging of pelvic lymph nodes in patients with prostate cancer: a meta-analysis. Clin Radiol 2008; 63:387–395 [DOI] [PubMed] [Google Scholar]
- 7.Hambrock T, Somford DM, Hoeks C, et al. Magnetic resonance imaging guided prostate biopsy in men with repeat negative biopsies and increased prostate specific antigen. J Urol 2010; 183:520–527 [DOI] [PubMed] [Google Scholar]
- 8.Rosenkrantz AB, Verma S, Choyke P, et al. Prostate magnetic resonance imaging and magnetic resonance imaging targeted biopsy in patients with a prior negative biopsy: a consensus statement by AUA and SAR. J Urol 2016; 196:1613–1618 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Panebianco V, Sciarra A, Lisi D, et al. Prostate cancer: 1HMRS-DCEMR at 3T versus [(18)F]choline PET/CT in the detection of local prostate cancer recurrence in men with biochemical progression after radical retropubic prostatectomy (RRP). Eur J Radiol 2012; 81:700–708 [DOI] [PubMed] [Google Scholar]
- 10.Haider MA, Chung P, Sweet J, et al. Dynamic contrast-enhanced magnetic resonance imaging for localization of recurrent prostate cancer after external beam radiotherapy. Int J Radiat Oncol Biol Phys 2008; 70:425–430 [DOI] [PubMed] [Google Scholar]
- 11.Barentsz JO, Richenberg J, Clements R, et al. ; European Society of Urogenital Radiology. ESUR prostate MR guidelines 2012. Eur Radiol 2012; 22:746–757 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Weinreb JC, Barentsz JO, Choyke PL, et al. PI-RADS Prostate Imaging-Reporting and Data System: 2015, version 2. Eur Urol 2016; 69:16–40 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Turkbey B, Rosenkrantz AB, Haider MA, et al. Prostate Imaging Reporting and Data System version 2.1: 2019 update of Prostate Imaging Reporting and Data System Version 2. Eur Urol 2019; 76:340–351 [DOI] [PubMed] [Google Scholar]
- 14.Greer MD, Brown AM, Shih JH, et al. Accuracy and agreement of PIRADSv2 for prostate cancer mpMRI: a multireader study. J Magn Reson Imaging 2017; 45:579–585 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rosenkrantz AB, Oto A, Turkbey B, Westphalen AC. Prostate Imaging Reporting and Data System (PI-RADS), version 2: a critical look. AJR 2016; 206:1179–1183 [DOI] [PubMed] [Google Scholar]
- 16.Sackett J, Choyke PL, Turkbey B. Prostate Imaging Reporting and Data System version 2 for MRI of prostate cancer: can we do better? AJR 2019; 212:1244–1252 [DOI] [PubMed] [Google Scholar]
- 17.Padhani AR, Barentsz J, Villeirs G, et al. ; PI-RADS Steering Committee. PI-RADS Steering Committee: the PI-RADS multiparametric MRI and MRI-directed biopsy pathway. Radiology 2019; 292:464–474 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.European Society of Radiology (ESR). Impact of artificial intelligence on radiology: a EuroAIM survey among members of the European Society of Radiology. Insights Imaging 2019; 10:105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rajiah P. Dual-energy computed tomography in thoracic imaging: current practices and utility survey of the Society of Thoracic Radiology. J Thorac Imaging 2020; 35:W43–W50 [DOI] [PubMed] [Google Scholar]
- 20.Spilseth B Ghai S, Patel NU, Taneja SS, Margolis DJ, Rosenkrantz AB. A comparison of radiologists’ and urologists’ opinion regarding prostate MRI reporting: results from a survey of specialty societies. AJR 2018; 210:101–107 [DOI] [PubMed] [Google Scholar]
- 21.Gupta RT, Spilseth B, Froemming AT. How and why a generation of radiologists must be trained to accurately interpret prostate mpMRI. Abdom Radiol (NY) 2016; 41:803–804 [DOI] [PubMed] [Google Scholar]
- 22.Bjurlin MA, Carroll PR, Eggener S, et al. Update of the standard operating procedure on the use of multiparametric magnetic resonance imaging for the diagnosis, staging and management of prostate cancer J Urol 2020; 203:706–712 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Agarwal HK, Mertan FV, Sankineni S, et al. Optimal high b-value for diffusion weighted MRI in diagnosing high risk prostate cancers in the peripheral zone. J Magn Reson Imaging 2017; 45:125–131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rosenkrantz AB, Parikh N, Kierans AS, et al. Prostate cancer detection using computed very high b-value diffusion-weighted imaging: how high should we go? Acad Radiol 2016; 23:704–711 [DOI] [PubMed] [Google Scholar]
- 25.Esses SJ, Taneja SS, Rosenkrantz AB. Imaging facilities’ adherence to PI-RADS v2 minimum technical standards for the performance of prostate MRI. Acad Radiol 2018; 25:188–195 [DOI] [PubMed] [Google Scholar]
- 26.Cuocolo R, Stanzione A, Ponsiglione A, et al. Prostate MRI technical parameters standardization: a systematic review on adherence to PI-RADSv2 acquisition protocol. Eur J Radiol 2019; 120:108662. [DOI] [PubMed] [Google Scholar]
- 27.Burn PR, Freeman SJ, Andreou A, Burns-Cox N, Persad R, Barrett T. A multicentre assessment of prostate MRI quality and compliance with UK and international standards. Clin Radiol 2019; 74:894.e19–894.e25 [DOI] [PubMed] [Google Scholar]
- 28.Tooker GM, Truong H, Pinto PA, Siddiqui MM. National survey of patterns employing targeted MRI/US guided prostate biopsy in the diagnosis and staging of prostate cancer. Curr Urol 2019; 12:97–103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Rosenkrantz AB, Lim RP, Haghighi M, Somberg MB, Babb JS, Taneja SS. Comparison of interreader reproducibility of the prostate imaging reporting and data system and Likert scales for evaluation of multiparametric prostate MRI. AJR 2013; 201:[web]W612–W618 [DOI] [PubMed] [Google Scholar]
- 30.Spilseth B, Margolis DJ, Ghai Sangeet, Patel NU, Rosenkrantz AB. Radiologist’s preferences regarding content of prostate MRI reports: a survey of the Society of Abdominal Radiology. Abdom Radiol 2018;43:1807–1812 [DOI] [PubMed] [Google Scholar]
- 31.Rosenkrantz AB, Ginocchio LA, Cornfeld D, et al. Interobserver reproducibility of the PI-RADS version 2 lexicon: a multicenter study of six experienced prostate radiologists. Radiology 2016; 280:793–804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Westphalen AC, McCulloch CE, Anaokar JM, et al. Variability of the positive predictive value of PI-RADS for prostate MRI across 26 centers: experience of the Society of Abdominal Radiology prostate cancer disease–focused panel. Radiology 2020; 296:76–84 [DOI] [PMC free article] [PubMed] [Google Scholar]