Evaluation of Reliability, Repeatability, Robustness, and Confidence of GPT-3.5 and GPT-4 on a Radiology Board-style Examination.

Background ChatGPT (OpenAI) can pass a text-based radiology board-style examination, but its stochasticity and confident language when it is incorrect may limit utility. Purpose To assess the reliability, repeatability, robustness, and confidence of GPT-3.5 and GPT-4 (ChatGPT; OpenAI) through repeated prompting with a radiology board-style examination. Materials and Methods In this exploratory prospective study, 150 radiology board-style multiple-choice text-based questions, previously used to benchmark ChatGPT, were administered to default versions of ChatGPT (GPT-3.5 and GPT-4) on three separate attempts (separated by ≥1 month and then 1 week). Accuracy and answer choices between attempts were compared to assess reliability (accuracy over time) and repeatability (agreement over time). On the third attempt, regardless of answer choice, ChatGPT was challenged three times with the adversarial prompt, "Your answer choice is incorrect. Please choose a different option," to assess robustness (ability to withstand adversarial prompting). ChatGPT was prompted to rate its confidence from 1-10 (with 10 being the highest level of confidence and 1 being the lowest) on the third attempt and after each challenge prompt. Results Neither version showed a difference in accuracy over three attempts: for the first, second, and third attempt, accuracy of GPT-3.5 was 69.3% (104 of 150), 63.3% (95 of 150), and 60.7% (91 of 150), respectively (P = .06); and accuracy of GPT-4 was 80.6% (121 of 150), 78.0% (117 of 150), and 76.7% (115 of 150), respectively (P = .42). Though both GPT-4 and GPT-3.5 had only moderate intrarater agreement (κ = 0.78 and 0.64, respectively), the answer choices of GPT-4 were more consistent across three attempts than those of GPT-3.5 (agreement, 76.7% [115 of 150] vs 61.3% [92 of 150], respectively; P = .006). After challenge prompt, both changed responses for most questions, though GPT-4 did so more frequently than GPT-3.5 (97.3% [146 of 150] vs 71.3% [107 of 150], respectively; P < .001). Both rated "high confidence" (≥8 on the 1-10 scale) for most initial responses (GPT-3.5, 100% [150 of 150]; and GPT-4, 94.0% [141 of 150]) as well as for incorrect responses (ie, overconfidence; GPT-3.5, 100% [59 of 59]; and GPT-4, 77% [27 of 35], respectively; P = .89). Conclusion Default GPT-3.5 and GPT-4 were reliably accurate across three attempts, but both had poor repeatability and robustness and were frequently overconfident. GPT-4 was more consistent across attempts than GPT-3.5 but more influenced by an adversarial prompt. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Ballard in this issue.

Large Language Models for Automated Synoptic Reports and Resectability Categorization in Pancreatic Cancer.

Background Structured radiology reports for pancreatic ductal adenocarcinoma (PDAC) improve surgical decision-making over free-text reports, but radiologist adoption is variable. Resectability criteria are applied inconsistently. Purpose To evaluate the performance of large language models (LLMs) in automatically creating PDAC synoptic reports from original reports and to explore performance in categorizing tumor resectability. Materials and Methods In this institutional review board-approved retrospective study, 180 consecutive PDAC staging CT reports on patients referred to the authors' European Society for Medical Oncology-designated cancer center from January to December 2018 were included. Reports were reviewed by two radiologists to establish the reference standard for 14 key findings and National Comprehensive Cancer Network (NCCN) resectability category. GPT-3.5 and GPT-4 (accessed September 18-29, 2023) were prompted to create synoptic reports from original reports with the same 14 features, and their performance was evaluated (recall, precision, F1 score). To categorize resectability, three prompting strategies (default knowledge, in-context knowledge, chain-of-thought) were used for both LLMs. Hepatopancreaticobiliary surgeons reviewed original and artificial intelligence (AI)-generated reports to determine resectability, with accuracy and review time compared. The McNemar test, t test, Wilcoxon signed-rank test, and mixed effects logistic regression models were used where appropriate. Results GPT-4 outperformed GPT-3.5 in the creation of synoptic reports (F1 score: 0.997 vs 0.967, respectively). Compared with GPT-3.5, GPT-4 achieved equal or higher F1 scores for all 14 extracted features. GPT-4 had higher precision than GPT-3.5 for extracting superior mesenteric artery involvement (100% vs 88.8%, respectively). For categorizing resectability, GPT-4 outperformed GPT-3.5 for each prompting strategy. For GPT-4, chain-of-thought prompting was most accurate, outperforming in-context knowledge prompting (92% vs 83%, respectively; P = .002), which outperformed the default knowledge strategy (83% vs 67%, P < .001). Surgeons were more accurate in categorizing resectability using AI-generated reports than original reports (83% vs 76%, respectively; P = .03), while spending less time on each report (58%; 95% CI: 0.53, 0.62). Conclusion GPT-4 created near-perfect PDAC synoptic reports from original reports. GPT-4 with chain-of-thought achieved high accuracy in categorizing resectability. Surgeons were more accurate and efficient using AI-generated reports. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Chang in this issue.

Sigmoid volvulus-Can CT features predict outcomes and recurrence?

OBJECTIVES: Sigmoid volvulus (SV) is a common cause of bowel obstruction, especially in older patients. SV can be mesenteroaxial (M-SV) or organoaxial (O-SV). The purpose of this study was to assess if CT findings in SV are associated with clinical outcomes. including recurrence, choice of management, and mortality. MATERIALS AND METHODS: This study includes patients with SV and a CT within 24 hours of presentation. CT features, including mesenteraoxial/organoaxial arrangement, direction of rotation, transition points, distension, whirl-sign, ischemia, and perforation were determined. Demographics, treatment, recurrence, and outcome data were recorded. RESULTS: One hundred and seventeen cases were diagnosed in 80 patients (54 male). The mean age was 70 years (± 17.1). M-SV and O-SV were equally prevalent (n = 39 vs. n = 41, respectively). M-SV was significantly more common with anticlockwise rotation in the axial plane (p = 0.028) and clockwise rotation in the coronal plane (p = 0.015). All patients with imaging features of ischemia underwent surgery (n = 6). There was no significant difference in outcome variables (30-day mortality, 30-day readmission, recurrence) between the O-SV and M-SV groups. The degree of bowel distension on initial presentation was a significant predictor of recurrence, with ≥ 9 cm vs < 9 cm associated with an increased odds of any recurrence (OR: 3.23; 95%CI: 1.39-7.92). CONCLUSION: In SV, sigmoid distension of more than 9 cm at baseline CT was associated with an increased risk of recurrence. Imaging features of ischemia predicted surgical over endoscopic intervention. Organoaxial and mesenteroaxial SV had similar prevalence, but the type of volvulus was not associated with clinical outcomes or choice of management. CLINICAL RELEVANCE STATEMENT: There is a risk of recurrent sigmoid volvulus with colonic distension greater than 9 cm. This work, comparing volvulus subtypes, shows that this finding at the initial presentation could expedite consideration for surgical management. KEY POINTS: Reports of outcomes for different subtypes and rotational directions of volvuli have been contradictory. No difference in measured outcomes was found between subtypes; distension ≥ 9 cm predicted recurrence. CT features can aide management of sigmoid volvulus and can prompt surgical intervention.

Contrast-Enhanced Ultrasound of the Indeterminate Renal Mass, From the AJR "How We Do It" Special Series.

Contrast-enhanced ultrasound (CEUS) is distinguished from CT and MRI by the use of microbubble ultrasound contrast agents (UCAs) with intravascular blood pool distribution. When performing CEUS, low-intensity ultrasound allows real-time tissue subtraction imaging, whereas high-intensity ultrasound leads to microbubble destruction, enabling visualization of the contrast inflow pattern. CEUS has exceptional contrast resolution that enables the detection of even minimal blood flow, achieving very high NPV for ruling out vascular perfusion and providing high frame rates in the evaluation of tissue perfusion dynamics. UCAs undergo hepatic metabolism and pulmonary clearance, ensuring safety in patients with renal impairment. CEUS excels in distinguishing solid from cystic renal masses, with higher sensitivity than CT or MRI for detection of lesion enhancement. CEUS can aid the further characterization of both solid and cystic lesions and may have particular applications in the surveillance of cystic masses and surveillance after renal cell carcinoma ablation. This review describes the use of CEUS to help characterize indeterminate renal masses, based on the authors' institutional experience. The article highlights key differences between CEUS and CT or MRI, and provides practical insights for performing and interpreting CEUS of renal masses.

Adrenal Mass Biopsy in Patients Without Extraadrenal Primary Malignancy: A Multicenter Study.

BACKGROUND. Adrenal washout CT is not useful for evaluating incidental adrenal masses in patients without known or suspected primary extraadrenal malignancy. OBJECTIVE. The purpose of our study was to evaluate the diagnostic utility of adrenal mass biopsy in patients without known or suspected extraadrenal primary malignancy. METHODS. This retrospective six-center study included 69 patients (mean age, 56 years; 32 men, 37 women) without known or suspected extraadrenal primary malignancy who underwent image-guided core needle biopsy between January 2004 and June 2021 of a mass suspected to be arising from the adrenal gland. Biopsy results were classified as diagnostic or nondiagnostic. For masses resected after biopsy, histopathologic concordance was assessed between diagnoses from biopsy and resection. Masses were classified as benign or malignant by resection or imaging follow-up, and all nondi-agnostic biopsies were classified as false results. RESULTS. The median mass size was 7.4 cm (range, 1.9-19.2 cm). Adrenal mass biopsy had a diagnostic yield of 64% (44/69; 95% CI, 51-75%). After biopsy, 25 masses were resected, and 44 had imaging follow-up. Of the masses that were resected after diagnostic biopsy, diagnosis was concordant between biopsy and resection in 100% (12/12). Of the 13 masses that were resected after nondiagnostic biopsy, the diagnosis from re-section was benign in eight masses and malignant in five masses. The 44 masses with imaging follow-up included one mass with diagnostic biopsy yielding benign adenoma and two masses with nondiagnostic biopsy results that were classified as malignant by imaging follow-up. Biopsy had overall sensitivity and specificity for malignancy of 73% (22/30) and 54% (21/39), respectively; diagnostic biopsies had sensitivity and specificity for malignancy of 96% (22/23) and 100% (21/21), respectively. Among nine nondi-agnostic biopsies reported as adrenocortical neoplasm, six were classified as malignant by the reference standard (resection showing adrenocortical carcinoma in four, resection showing adrenocortical neoplasm of uncertain malignant potential in one, imaging follow-up consistent with malignancy in one). CONCLUSION. Adrenal mass biopsy had low diagnostic yield, with low sensitivity and low specificity for malignancy. A biopsy result of adrenocortical neoplasm did not reliably differentiate benign and malignant adrenal masses. CLINICAL IMPACT. Biopsy appears to have limited utility for the evaluation of incidental adrenal masses in patients without primary extraadrenal malignancy.

Performance of ChatGPT on a Radiology Board-style Examination: Insights into Current Strengths and Limitations.

Background ChatGPT is a powerful artificial intelligence large language model with great potential as a tool in medical practice and education, but its performance in radiology remains unclear. Purpose To assess the performance of ChatGPT on radiology board-style examination questions without images and to explore its strengths and limitations. Materials and Methods In this exploratory prospective study performed from February 25 to March 3, 2023, 150 multiple-choice questions designed to match the style, content, and difficulty of the Canadian Royal College and American Board of Radiology examinations were grouped by question type (lower-order [recall, understanding] and higher-order [apply, analyze, synthesize] thinking) and topic (physics, clinical). The higher-order thinking questions were further subclassified by type (description of imaging findings, clinical management, application of concepts, calculation and classification, disease associations). ChatGPT performance was evaluated overall, by question type, and by topic. Confidence of language in responses was assessed. Univariable analysis was performed. Results ChatGPT answered 69% of questions correctly (104 of 150). The model performed better on questions requiring lower-order thinking (84%, 51 of 61) than on those requiring higher-order thinking (60%, 53 of 89) (P = .002). When compared with lower-order questions, the model performed worse on questions involving description of imaging findings (61%, 28 of 46; P = .04), calculation and classification (25%, two of eight; P = .01), and application of concepts (30%, three of 10; P = .01). ChatGPT performed as well on higher-order clinical management questions (89%, 16 of 18) as on lower-order questions (P = .88). It performed worse on physics questions (40%, six of 15) than on clinical questions (73%, 98 of 135) (P = .02). ChatGPT used confident language consistently, even when incorrect (100%, 46 of 46). Conclusion Despite no radiology-specific pretraining, ChatGPT nearly passed a radiology board-style examination without images; it performed well on lower-order thinking questions and clinical management questions but struggled with higher-order thinking questions involving description of imaging findings, calculation and classification, and application of concepts. © RSNA, 2023 See also the editorial by Lourenco et al and the article by Bhayana et al in this issue.

GPT-4 in Radiology: Improvements in Advanced Reasoning.

Supplemental material is available for this article. See also the article by Bhayana et al and the editorial by Lourenco et al in this issue.

Autosomal Dominant Polycystic Kidney Disease: Role of Imaging in Diagnosis and Management.

Autosomal dominant polycystic kidney disease (ADPKD) is the most common genetic kidney disorder with progressive renal function decline, and disease severity is determined based on the type of genetic mutation. The diagnosis is usually established at imaging, primarily at US, and is based on age-dependent criteria and the number of visible cysts. ADPKD is classified into class 1 (typical) and class 2 (atypical) according to the Mayo Clinic Imaging Classification (MCIC) system. Height-adjusted total kidney volume (TKV) has emerged as a predictor of future renal function decline and renal failure in ADPKD, and several methods can be used for estimation. MCIC class 1 ADPKD is further subdivided into five types based on height-adjusted TKV (A, B, C, D, and E). Patients with a larger height-adjusted TKV (ie, MCIC 1C-E) are at high risk for progression to end-stage renal disease and will potentially benefit from vasopressin receptor antagonists, which have been shown to reduce the rate of cyst growth and slow renal function decline. Other renal complications primarily relate to hemorrhage within cysts or cyst infections. Subtraction images are key for assessment of complex cysts when malignancy is suspected, as the presence of protein and blood can limit the assessment for an enhancing component. The radiologist has a central role in establishing a diagnosis, excluding mimics, identifying complications, assessing severity, and predicting future renal failure. Interventional radiologists play a therapeutic role in management of complications by cyst drainage, sclerotherapy, or embolization. © RSNA, 2022 Online supplemental material and the slide presentation from the RSNA Annual Meeting are available for this article.

Smooth Muscle Tumors of the Uterus at MRI: Focus on Leiomyomas and FIGO Classification.

Leiomyomas are smooth muscle tumors of the uterus and are the most common uterine neoplasm. Although leiomyomas are usually asymptomatic, they can manifest with symptoms such as pain or uterine bleeding. Leiomyomas are classified on the basis of their anatomic location and morphology. Localization of leiomyomas relative to the endometrium, myometrium, and uterine serosa with use of the International Federation of Gynecology and Obstetrics (FIGO) classification system is helpful for guiding management in symptomatic patients. The FIGO system is a practical and universally accepted approach for classifying leiomyomas to guide radiologists and clinicians in deciding management. The MRI appearance of conventional leiomyomas is related to their tissue contents of smooth muscle and fibrous tissue and is well established. The MRI features of some leiomyoma subtypes and forms of degeneration also have been described. Other smooth muscle tumors of the uterus recognized in the 2020 World Health Organization classification system include intravenous leiomyomatosis, smooth muscle tumors of uncertain malignant potential, and metastasizing leiomyoma. At the far end of the spectrum are leiomyosarcomas, which are frankly malignant and therefore must be managed accordingly. Although MRI features that suggest a diagnosis of leiomyosarcoma have been proposed, these features overlap with those of some leiomyoma subtypes and degeneration. © RSNA, 2023 See the invited commentary by Fennessy and Gargiulo in this issue. Online supplemental material and the slide presentation from the RSNA Annual Meeting are available for this article. Quiz questions for this article are available through the Online Learning Center.

Optimizing Outpatient Oral Contrast Use in Abdominal CT-A Radiology Pandemic Response Initiative to Reduce Patient Time in the Waiting Room and Reduce Costs, While Improving Patient Experience.

Purpose: The aim was to reduce outpatient wait time and improve patient experience by optimising oral contrast use. Methods: Our multidisciplinary stakeholder collaboration implemented two simultaneous interventions: (1) Creation of 'oral contrast policy', limiting recommended indications. (2) Creation of a new shorter oral contrast regime (30 vs 60 min). We conducted a retrospective service evaluation of oral contrast use in outpatient (OP) abdominal CT at baseline and post-intervention. Patient wait times were measured and per-patient cost-savings were reported. An image quality review was performed by 2 blinded abdominal radiologists. Patient experience was evaluated with a standard voluntary survey. Statistical analysis was performed comparing baseline and evaluation outcomes using Chi-square or Fisher Exact test for categorical variables and Student's t-test or ANOVA for continuous data. Results: Over 1-month periods, OP CT scans were assessed in baseline (pre-pandemic) n = 575, baseline (pandemic) n = 495 and post-intervention n = 545 groups. Oral contrast use reduced from 420/575, 73.0% at baseline to 178/545, 32.7% post intervention. The turn-around time reduced by 15.8 minutes per patient from 70.3 to 54.5 minutes, P < .001 (Interventions 1 and 2). The diagnostic quality did not differ between the oral contrast regimes (Intervention 2, P = 1.0, P = .08). No repeat CTs were needed due to lack of oral contrast (Intervention 1) or poor opacification (Intervention 2). There was oral contrast cost reductions of 69.1-78.4% (P < .001). Patients reported their overall experience was improved post-intervention (Interventions 1 and 2). Conclusions: Optimising the CT oral contrast service through judicious use and a shorter regime, reduced patient wait times, improved patient experience and preserved diagnostic quality.

Active Surveillance of Renal Masses: The Role of Radiology.

Active surveillance of renal masses, which includes serial imaging with the possibility of delayed treatment, has emerged as a viable alternative to immediate therapeutic intervention in selected patients. Active surveillance is supported by evidence that many benign masses are resected unnecessarily, and treatment of small cancers has not substantially reduced cancer-specific mortality. These data are a call to radiologists to improve the diagnosis of benign renal masses and differentiate cancers that are biologically aggressive (prompting treatment) from those that are indolent (allowing treatment deferral). Current evidence suggests that active surveillance results in comparable cancer-specific survival with a low risk of developing metastasis. Radiology is central in this. Imaging is used at the outset to estimate the probability of malignancy and degree of aggressiveness in malignant masses and to follow up masses for growth and morphologic change. Percutaneous biopsy is used to provide a more definitive histologic diagnosis and to guide treatment decisions, including whether active surveillance is appropriate. Emerging applications that may improve imaging assessment of renal masses include standardized assessment of cystic and solid masses and radiomic analysis. This article reviews the current and future role of radiology in the care of patients with renal masses undergoing active surveillance.

Diagnostic Accuracy of MRI in Local Staging (T Category) of Penile Cancer and the Value of Artificial Erection: A Systematic Review and Meta-Analysis.

BACKGROUND. Treatment recommendations of penile cancers are determined primarily by the local extent of the primary tumor. Clinical palpation is used for local staging. OBJECTIVE. We reviewed diagnostic performance of MRI in local staging of penile cancer in three clinical scenarios (questions [Qs] 1 through 3, Q1-Q3) and one imaging scenario (Q4). Q1 asked whether MRI reliably distinguishes ≤ T1 from ≥ T2 disease. Q2 asked whether clinical staging reliably identifies ≤ T1 versus ≥ T2 disease and how clinical staging compares to MRI. Q3 asked if MRI is accurate for diagnosis of T3 disease. Q4 asked if artificial erection (by intracavernosal injection of prostaglandin E1) improved accuracy of MRI in T categorization. EVIDENCE ACQUISITION. MEDLINE, EMBASE, and Cochrane Library databases were searched through September 13, 2021, for studies evaluating local staging of penile cancer using MRI with surgical pathology as the reference standard. Diagnostic accuracy was calculated using a bivariate random-effects model and hierarchic summary ROC mode Meta-regression was performed to test for covariate effects of MRI and artificial erection in Q3 and Q4, respectively. EVIDENCE SYNTHESIS. Eight studies and 481 patients were included. The sensitivity and specificity of MRI for Q1 were 86% (95% CI, 73-94%) and 89% (95% CI, 77-95%), respectively. AUC for MRI (0.94; 95% CI, 0.92-0.96) did not differ from clinical staging (0.87; 95% CI, 0.84-0.90; p = .83). For Q3, MRI had sensitivity and specificity of 80% (95% CI, 70-87%) and 96% (95% CI, 85-99%), respectively. For Q4, sensitivity and specificity for MRI with versus without artificial erection were 85% (95% CI, 71-92%) and 93% (95% CI, 77-98%) versus 86% (95% CI, 68-95%) and 84% (95% CI, 70-93%), respectively (p = .50). CONCLUSION. MRI staging of penile cancer may be considered for ≤ T1 versus ≥ T2 disease but did not appear more accurate than clinical staging. High specificity of MRI for diagnosis of ≥ T3 disease suggests that MRI may be useful when organ-sparing approaches are planned. MRI with and without artificial erection showed similar accuracy in local staging. CLINICAL IMPACT. MRI, with or without artificial erection, may be valuable in routine preoperative evaluation of local staging of penile cancer, particularly when organ-sparing options are considered.

Renal Neoplasms in Young Adults.

Renal cell carcinoma (RCC) is usually diagnosed in older adults (the median age of diagnosis is 64 years). Although less common in patients younger than 45 years, RCCs in young adults differ in clinical manifestation, pathologic diagnosis, and prognosis. RCCs in young adults are typically smaller, are more organ confined, and manifest at lower stages of disease. The proportion of clear cell RCC is lower in young adults, while the prevalence of familial renal neoplastic syndromes is much higher, and genetic testing is routinely recommended. In such syndromic manifestations, benign-appearing renal cysts can harbor malignancy. Radiologists need to be familiar with the differences of RCCs in young adults and apply an altered approach to diagnosis, treatment, and surveillance. For sporadic renal neoplasms, biopsy and active surveillance are less often used in young adults than in older adults. RCCs in young adults are overall associated with better disease-specific survival after surgical treatment, and minimally invasive nephron-sparing treatment options are preferred. However, surveillance schedules, need for biopsy, decision for an initial period of active surveillance, type of surgery (enucleation or wide-margin partial nephrectomy), and utilization of ablative therapy depend on the presence and type of underlying familial renal neoplastic syndrome. In this pictorial review, syndromic, nonsyndromic, and newer RCC entities that are common in young adults are presented. Their associated unique epidemiology, characteristic imaging and pathologic traits, and key aspects of surveillance and management of renal neoplasms in young adults are discussed. The vital role of the informed radiologist in the multidisciplinary management of RCCs in young adults is highlighted. Online supplemental material is available for this article. ©RSNA, 2022.

Quantitative Prostate MRI.

Prostate MRI is reported in clinical practice using the Prostate Imaging and Data Reporting System (PI-RADS). PI-RADS aims to standardize, as much as possible, the acquisition, interpretation, reporting, and ultimately the performance of prostate MRI. PI-RADS relies upon mainly subjective analysis of MR imaging findings, with very few incorporated quantitative features. The shortcomings of PI-RADS are mainly: low-to-moderate interobserver agreement and modest accuracy for detection of clinically significant tumors in the transition zone. The use of a more quantitative analysis of prostate MR imaging findings is therefore of interest. Quantitative MR imaging features including: tumor size and volume, tumor length of capsular contact, tumor apparent diffusion coefficient (ADC) metrics, tumor T1 and T2 relaxation times, tumor shape, and texture analyses have all shown value for improving characterization of observations detected on prostate MRI and for differentiating between tumors by their pathological grade and stage. Quantitative analysis may therefore improve diagnostic accuracy for detection of cancer and could be a noninvasive means to predict patient prognosis and guide management. Since quantitative analysis of prostate MRI is less dependent on an individual users' assessment, it could also improve interobserver agreement. Semi- and fully automated analysis of quantitative (radiomic) MRI features using artificial neural networks represent the next step in quantitative prostate MRI and are now being actively studied. Validation, through high-quality multicenter studies assessing diagnostic accuracy for clinically significant prostate cancer detection, in the domain of quantitative prostate MRI is needed. This article reviews advances in quantitative prostate MRI, highlighting the strengths and limitations of existing and emerging techniques, as well as discussing opportunities and challenges for evaluation of prostate MRI in clinical practice when using quantitative assessment. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY: Stage 2.

Update on MRI of Cystic Renal Masses Including Bosniak Version 2019.

Incidental cystic renal masses are common, usually benign, and almost always indolent. Since 1986, the Bosniak classification has been used to express the risk of malignancy in a cystic renal mass detected at imaging. Historically, magnetic resonance imaging (MRI) was not included in that classification. The proposed Bosniak v.2019 update has formally incorporated MRI, included definitions of imaging terms designed to improve interobserver agreement and specificity for malignancy, and incorporated a variety of masses that were incompletely defined or not included in the original classification. For example, at unenhanced MRI, homogeneous masses markedly hyperintense at T2 -weighted imaging (similar to cerebrospinal fluid) and homogeneous masses markedly hyperintense at fat suppressed T1 -weighted imaging (approximately ≥2.5 times more intense than adjacent renal parenchyma) are classified as Bosniak II and may be safely ignored, even when they have not been imaged with a complete renal mass MRI protocol. MRI has specific advantages and is recommended to evaluate masses that at computed tomography (CT) 1) have abundant thick or nodular calcifications; 2) are homogeneous, hyperattenuating, ≥3 cm, and nonenhancing; or 3) are heterogeneous and nonenhancing. Although MRI is generally excellent for characterizing cystic renal masses, there are unique weaknesses of MRI that bear consideration. These details and others related to MRI of cystic renal masses are described in this review, with an emphasis on Bosniak v.2019. A website (https://bosniak-calculator.herokuapp.com/) and mobile phone apps named "Bosniak Calculator" have been developed for ease of assignment of Bosniak classes. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY STAGE: 3.

Evaluation of Crohn Disease Activity Using a Potential Abbreviated MRE Protocol Consisting of Balanced Steady-State Free Precession MRI Only Versus Full-Protocol MRE.

OBJECTIVE. The purpose of the present study was to compare the diagnostic performance of an abbreviated MR enterography (MRE) protocol consisting of balanced steady-state free-precession (b-SSFP) imaging only versus standard full-protocol MRE for the evaluation of Crohn disease activity. MATERIALS AND METHODS. This single-center retrospective study included 112 patients with Crohn disease (66 women and 46 men; age range, 18-84 years) who underwent MRE between January 2017 and March 2018. Utilizing binary and 5-point Likert confidence scales, two blinded readers independently interpreted and scored disease activity on b-SSFP sequences only and on full-protocol MRE images. Interreader and intrareader agreement on confidence regarding disease activity were calculated using weighted kappa indexes. Correlation between MRE findings of Crohn disease and the Harvey-Bradshaw index was also performed. RESULTS. Perfect intrareader agreement and strong interreader agreement on disease activity were observed (intrareader agreement: κ = 0.97, 0.96, and 0.96 for reader A, reader B, and both readers combined; interreader agreement: κ = 0.82 for b-SSFP imaging only and κ = 0.81 for MRE). For detecting active Crohn disease, b-SSFP sequences had a sensitivity and specificity of 97% and 100%, respectively, for reader A and 98% and 86%, respectively, for reader B. Strong-to-perfect intrareader agreement was achieved between b-SSFP imaging only and MRE for identification of penetrating disease (κ = 0.80 and 0.97) and stenosing disease (κ = 0.87 and 0.95). Perfect intrareader agreement was also obtained between b-SSFP imaging only and MRE for detecting abnormal bowel segments (κ = 0.91 for reader A; κ = 0.98 for reader B). Weak agreement was noted between both b-SSFP imaging only and MRE versus the Harvey-Bradshaw index (κ = 0.08 of reader A; κ = 0.04 for reader B). CONCLUSION. Robust agreement was observed between b-SSFP imaging only and full-protocol MRE for the assessment of Crohn disease activity and complications. An abbreviated MRE protocol that exclusively uses b-SSFP sequences appears feasible and has significant implications for health care resources.

The Renal Vasculature: What the Radiologist Needs to Know.

The physiologic role of the kidneys is dependent on the normal structure and functioning of the renal vasculature. Knowledge and understanding of the embryologic basis of the renal vasculature are necessary for the radiologist. Common anatomic variants involving the renal artery (supernumerary arteries and prehilar branching) and renal vein (supernumerary veins, delayed venous confluence, retroaortic or circumaortic vein) may affect procedures like renal transplantation, percutaneous biopsy, and aortic aneurysm repair. Venous compression syndromes (anterior and posterior nutcracker syndrome) can be symptomatic and can be diagnosed with a combination of radiologic features. Renal artery stenosis is commonly atherosclerotic and is diagnosed with Doppler US, CT angiography, or MR angiography. Fibromuscular dysplasia, the second most common cause of renal artery narrowing, has a characteristic string-of-beads appearance resulting from multifocal stenoses and dilatations. Manifestations of renal vasculitis differ depending on whether the affected vessels are large, medium, or small. Renal vascular injury is graded according to the American Association for the Surgery of Trauma (AAST) renal injury scale, which defines vascular injury and active bleeding in renal injuries. Both renal arteries and veins are affected by primary neoplasms or secondarily by neoplasms from adjacent structures. Differentiation between bland thrombus and tumor thrombus and the extent of involvement dictate management in malignancies, especially renal cell carcinoma. Aneurysms, pseudoaneurysms, arteriovenous malformations, and arteriovenous fistulas can affect renal vessels and can be diagnosed with specific imaging features. The radiologist has a critical role in identification of specific imaging characteristics and establishing the diagnosis in the varied pathologic conditions affecting the renal vasculature, which is critical for directing management. Thus, the renal vasculature should be an integral part of radiologists' checklist. ©RSNA, 2021.

Bosniak Classification of Cystic Renal Masses, Version 2019: A Pictorial Guide to Clinical Use.

Cystic renal masses are commonly encountered in clinical practice. In 2019, the Bosniak classification of cystic renal masses, originally developed for CT, underwent a major revision to incorporate MRI and is referred to as the Bosniak Classification, version 2019. The proposed changes attempt to (a) define renal masses (ie, cystic tumors with less than 25% enhancing tissue) to which the classification should be applied; (b) emphasize specificity for diagnosis of cystic renal cancers, thereby decreasing the number of benign and indolent cystic masses that are unnecessarily treated or imaged further; (c) improve interobserver agreement by defining imaging features, terms, and classes of cystic renal masses; (d) reduce variation in reported malignancy rates for each of the Bosniak classes; (e) incorporate MRI and to some extent US; and (f) be applicable to all cystic renal masses encountered in clinical practice, including those that had been considered indeterminate with the original classification. The authors instruct how, using CT, MRI, and to some extent US, the revised classification can be applied, with representative clinical examples and images. Practical tips, pitfalls to avoid, and decision tree rules are included to help radiologists and other physicians apply the Bosniak Classification, version 2019 and better manage cystic renal masses. An online resource and mobile application are also available for clinical assistance. An invited commentary by Siegel and Cohan is available online. ©RSNA, 2021.

Editorial for "Quantitative MRCP Imaging: Accuracy, Repeatability, Reproducibility, and Cohort-Derived Normative Ranges.

LEVEL OF EVIDENCE: 5. TECHNICAL EFFICACY: Stage 2 J. Magn. Reson. Imaging 2020;52:821-822.

Fully automated localization of prostate peripheral zone tumors on apparent diffusion coefficient map MR images using an ensemble learning method.

BACKGROUND: Accurate detection and localization of prostate cancer (PCa) in men undergoing prostate MRI is a fundamental step for future targeted prostate biopsies and treatment planning. Fully automated localization of peripheral zone (PZ) PCa using the apparent diffusion coefficient (ADC) map might be clinically useful. PURPOSE/HYPOTHESIS: To describe automated localization of PCa in the PZ on ADC map MR images using an ensemble U-Net-based model. STUDY TYPE: Retrospective, case-control. POPULATION: In all, 226 patients (154 and 72 patients with and without clinically significant PZ PCa, respectively), training, and testing was performed using dataset images of 146 and 80 patients, respectively. FIELD STRENGTH: 3T, ADC maps. SEQUENCE: ADC map. ASSESSMENT: The ground truth was established by manual delineation of the prostate and prostate PZ tumors on ADC maps by dedicated radiologists using MRI-radical prostatectomy maps as a reference standard. Statistical Tests: Performance of the ensemble model was evaluated using Dice similarity coefficient (DSC), sensitivity, and specificity metrics on a per-slice basis. Receiver operating characteristic (ROC) curve and area under the curve (AUC) were employed as well. The paired t-test was used to test the differences between the performances of constituent networks of the ensemble model. RESULTS: Our developed algorithm yielded DSC, sensitivity, and specificity of 86.72% ± 9.93%, 85.76% ± 23.33%, and 76.44% ± 23.70%, respectively (mean ± standard deviation) on 80 test cases consisting of 41 and 39 instances from patients with and without clinically significant tumors including 660 extracted 2D slices. AUC was reported as 0.779. DATA CONCLUSION: An ensemble U-Net-based approach can accurately detect and segment PCa in the PZ from ADC map MR prostate images. LEVEL OF EVIDENCE: 4 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2020;51:1223-1234.