Evaluation of ChatGPT-Generated Educational Patient Pamphlets for Common Interventional Radiology Procedures.

RATIONALE AND OBJECTIVES: This study aimed to evaluate the accuracy and reliability of educational patient pamphlets created by ChatGPT, a large language model, for common interventional radiology (IR) procedures. METHODS AND MATERIALS: Twenty frequently performed IR procedures were selected, and five users were tasked to independently request ChatGPT to generate educational patient pamphlets for each procedure using identical commands. Subsequently, two independent radiologists assessed the content, quality, and accuracy of the pamphlets. The review focused on identifying potential errors, inaccuracies, the consistency of pamphlets. RESULTS: In a thorough analysis of the education pamphlets, we identified shortcomings in 30% (30/100) of pamphlets, with a total of 34 specific inaccuracies, including missing information about sedation for the procedure (10/34), inaccuracies related to specific procedural-related complications (8/34). A key-word co-occurrence network showed consistent themes within each group of pamphlets, while a line-by-line comparison at the level of users and across different procedures showed statistically significant inconsistencies (P < 0.001). CONCLUSION: ChatGPT-generated education pamphlets demonstrated potential clinical relevance and fairly consistent terminology; however, the pamphlets were not entirely accurate and exhibited some shortcomings and inter-user structural variabilities. To ensure patient safety, future improvements and refinements in large language models are warranted, while maintaining human supervision and expert validation.

ChatGPT in radiology: A systematic review of performance, pitfalls, and future perspectives.

PURPOSE: The purpose of this study was to systematically review the reported performances of ChatGPT, identify potential limitations, and explore future directions for its integration, optimization, and ethical considerations in radiology applications. MATERIALS AND METHODS: After a comprehensive review of PubMed, Web of Science, Embase, and Google Scholar databases, a cohort of published studies was identified up to January 1, 2024, utilizing ChatGPT for clinical radiology applications. RESULTS: Out of 861 studies derived, 44 studies evaluated the performance of ChatGPT; among these, 37 (37/44; 84.1%) demonstrated high performance, and seven (7/44; 15.9%) indicated it had a lower performance in providing information on diagnosis and clinical decision support (6/44; 13.6%) and patient communication and educational content (1/44; 2.3%). Twenty-four (24/44; 54.5%) studies reported the proportion of ChatGPT's performance. Among these, 19 (19/24; 79.2%) studies recorded a median accuracy of 70.5%, and in five (5/24; 20.8%) studies, there was a median agreement of 83.6% between ChatGPT outcomes and reference standards [radiologists' decision or guidelines], generally confirming ChatGPT's high accuracy in these studies. Eleven studies compared two recent ChatGPT versions, and in ten (10/11; 90.9%), ChatGPTv4 outperformed v3.5, showing notable enhancements in addressing higher-order thinking questions, better comprehension of radiology terms, and improved accuracy in describing images. Risks and concerns about using ChatGPT included biased responses, limited originality, and the potential for inaccurate information leading to misinformation, hallucinations, improper citations and fake references, cybersecurity vulnerabilities, and patient privacy risks. CONCLUSION: Although ChatGPT's effectiveness has been shown in 84.1% of radiology studies, there are still multiple pitfalls and limitations to address. It is too soon to confirm its complete proficiency and accuracy, and more extensive multicenter studies utilizing diverse datasets and pre-training techniques are required to verify ChatGPT's role in radiology.

Ferumoxytol-Enhanced Cardiac Cine MRI Reconstruction Using a Variable-Splitting Spatiotemporal Network.

BACKGROUND: Balanced steady-state free precession (bSSFP) imaging is commonly used in cardiac cine MRI but prone to image artifacts. Ferumoxytol-enhanced (FE) gradient echo (GRE) has been proposed as an alternative. Utilizing the abundance of bSSFP images to develop a computationally efficient network that is applicable to FE GRE cine would benefit future network development. PURPOSE: To develop a variable-splitting spatiotemporal network (VSNet) for image reconstruction, trained on bSSFP cine images and applicable to FE GRE cine images. STUDY TYPE: Retrospective and prospective. SUBJECTS: 41 patients (26 female, 53 ± 19 y/o) for network training, 31 patients (19 female, 49 ± 17 y/o) and 5 healthy subjects (5 female, 30 ± 7 y/o) for testing. FIELD STRENGTH/SEQUENCE: 1.5T and 3T, bSSFP and GRE. ASSESSMENT: VSNet was compared to VSNet with total variation loss, compressed sensing and low rank methods for 14× accelerated data. The GRAPPA×2/×3 images served as the reference. Peak signal-to-noise-ratio (PSNR), structural similarity index (SSIM), left ventricular (LV) and right ventricular (RV) end-diastolic volume (EDV), end-systolic volume (ESV), and ejection fraction (EF) were measured. Qualitative image ranking and scoring were independently performed by three readers. Latent scores were calculated based on scores of each method relative to the reference. STATISTICS: Linear mixed-effects regression, Tukey method, Fleiss' Kappa, Bland-Altman analysis, and Bayesian categorical cumulative probit model. A P-value <0.05 was considered statistically significant. RESULTS: VSNet achieved significantly higher PSNR (32.7 ± 0.2), SSIM (0.880 ± 0.004), rank (2.14 ± 0.06), and latent scores (-1.72 ± 0.22) compared to other methods (rank >2.90, latent score < -2.63). Fleiss' Kappa was 0.52 for scoring and 0.61 for ranking. VSNet showed no significantly different LV and RV ESV (P = 0.938) and EF (P = 0.143) measurements, but statistically significant different (2.62 mL) EDV measurements compared to the reference. CONCLUSION: VSNet produced the highest image quality and the most accurate functional measurements for FE GRE cine images among the tested 14× accelerated reconstruction methods. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 1.

How AI Responds to Common Lung Cancer Questions: ChatGPT vs Google Bard.

Background The recent release of large language models (LLMs) for public use, such as ChatGPT and Google Bard, has opened up a multitude of potential benefits as well as challenges. Purpose To evaluate and compare the accuracy and consistency of responses generated by publicly available ChatGPT-3.5 and Google Bard to non-expert questions related to lung cancer prevention, screening, and terminology commonly used in radiology reports based on the recommendation of Lung Imaging Reporting and Data System (Lung-RADS) v2022 from American College of Radiology and Fleischner society. Materials and Methods Forty of the exact same questions were created and presented to ChatGPT-3.5 and Google Bard experimental version as well as Bing and Google search engines by three different authors of this paper. Each answer was reviewed by two radiologists for accuracy. Responses were scored as correct, partially correct, incorrect, or unanswered. Consistency was also evaluated among the answers. Here, consistency was defined as the agreement between the three answers provided by ChatGPT-3.5, Google Bard experimental version, Bing, and Google search engines regardless of whether the concept conveyed was correct or incorrect. The accuracy among different tools were evaluated using Stata. Results ChatGPT-3.5 answered 120 questions with 85 (70.8%) correct, 14 (11.7%) partially correct, and 21 (17.5%) incorrect. Google Bard did not answer 23 (19.1%) questions. Among the 97 questions answered by Google Bard, 62 (51.7%) were correct, 11 (9.2%) were partially correct, and 24 (20%) were incorrect. Bing answered 120 questions with 74 (61.7%) correct, 13 (10.8%) partially correct, and 33 (27.5%) incorrect. Google search engine answered 120 questions with 66 (55%) correct, 27 (22.5%) partially correct, and 27 (22.5%) incorrect. The ChatGPT-3.5 is more likely to provide correct or partially answer than Google Bard, approximately by 1.5 folds (OR = 1.55, P = 0.004). ChatGPT-3.5 and Google search engine were more likely to be consistent than Google Bard by approximately 7 and 29 folds (OR = 6.65, P = 0.002 for ChatGPT and OR = 28.83, P = 0.002 for Google search engine, respectively). Conclusion Although ChatGPT-3.5 had a higher accuracy in comparison with the other tools, neither ChatGPT nor Google Bard, Bing and Google search engines answered all questions correctly and with 100% consistency.

Cardiac Outpouchings: Practical Approach to Normal Variants and Pathologic Conditions at CT and MRI.

Numerous entities, both structural and pathologic, can manifest as a contrast material- or blood-filled cardiac outpouching at imaging. These outpouchings often resemble one another and are frequently unfamiliar to imagers and clinicians, creating uncertainty when detected. Furthermore, the diagnostic criteria for conditions such as hernia, aneurysm, pseudoaneurysm, and diverticulum have not been consistently applied in studies and reports cited in the literature describing these outpouchings, adding to the confusion among general and cardiothoracic imagers. Pouches and outpouchings are commonly found incidentally on thoracic and abdominal CT scans obtained for other reasons. Many pouches and outpouchings can be confidently diagnosed or ignored at routine imaging, whereas others may require further evaluation with electrocardiographically gated CT, cardiac MRI, or echocardiography for a more definitive diagnosis. It is easiest to group and diagnose these entities on the basis of their cardiac chamber location or their involvement with the interatrial and interventricular septa. Ancillary features, such as motion, morphology, neck and body size, presence or absence of thrombus, and late gadolinium enhancement characteristics, are important in reaching a correct diagnosis. The aim of this article is to provide a practical guide to pouches and outpouchings of the heart. Each entity is defined according to its cause, imaging characteristics, clinical significance, and relevant associated findings. Mimics of cardiac pouches and outpouchings such as the Bachmann bundle, atrial veins, and thebesian vessels also are briefly discussed. Quiz questions for this article are available in the supplemental material. ©RSNA, 2023.

Contrast-enhanced MR Angiography without Gadolinium-based Contrast Material: Clinical Applications Using Ferumoxytol.

Vascular imaging can be challenging because of the wide variability of contrast dynamics in different vascular territories and potential safety concerns in patients with renal insufficiency or allergies. Off-label diagnostic use of ferumoxytol, a superparamagnetic iron nanoparticle approved for therapy, is a promising alternative to gadolinium-based contrast agents for MR angiography (MRA). Ferumoxytol has exhibited a reassuring safety profile when used within the dose range recommended for diagnostic imaging. Because of its prolonged and stable intravascular residence, ferumoxytol can be used in its steady-state distribution for a wide variety of imaging indications, including some where conventional MRA is unreliable. In this article, authors discuss some of the major vascular applications of ferumoxytol and highlight how it may be used to provide highly diagnostic images and improve the quality, workflow, and reliability of vascular imaging. Keywords: MR Angiography, MRI Contrast Agent, Cardiac, Vascular © RSNA, 2022.

Automatic segmentation of peripheral arteries and veins in ferumoxytol-enhanced MR angiography.

PURPOSE: To automate the segmentation of the peripheral arteries and veins in the lower extremities based on ferumoxytol-enhanced MR angiography (FE-MRA). METHODS: Our automated pipeline has 2 sequential stages. In the first stage, we used a 3D U-Net with local attention gates, which was trained based on a combination of the Focal Tversky loss with region mutual loss under a deep supervision mechanism to segment the vasculature from the high-resolution FE-MRA datasets. In the second stage, we used time-resolved images to separate the arteries from the veins. Because the ultimate segmentation quality of the arteries and veins relies on the performance of the first stage, we thoroughly evaluated the different aspects of the segmentation network and compared its performance in blood vessel segmentation with currently accepted state-of-the-art networks, including Volumetric-Net, DeepVesselNet-FCN, and Uception. RESULTS: We achieved a competitive F1 = 0.8087 and recall = 0.8410 for blood vessel segmentation compared with F1 = (0.7604, 0.7573, 0.7651) and recall = (0.7791, 0.7570, 0.7774) obtained with Volumetric-Net, DeepVesselNet-FCN, and Uception. For the artery and vein separation stage, we achieved F1 = (0.8274/0.7863) in the calf region, which is the most challenging region in peripheral arteries and veins segmentation. CONCLUSION: Our pipeline is capable of fully automatic vessel segmentation based on FE-MRA without need for human interaction in <4 min. This method improves upon manual segmentation by radiologists, which routinely takes several hours.

Temporally aware volumetric generative adversarial network-based MR image reconstruction with simultaneous respiratory motion compensation: Initial feasibility in 3D dynamic cine cardiac MRI.

PURPOSE: Develop a novel three-dimensional (3D) generative adversarial network (GAN)-based technique for simultaneous image reconstruction and respiratory motion compensation of 4D MRI. Our goal was to enable high-acceleration factors 10.7X-15.8X, while maintaining robust and diagnostic image quality superior to state-of-the-art self-gating (SG) compressed sensing wavelet (CS-WV) reconstruction at lower acceleration factors 3.5X-7.9X. METHODS: Our GAN was trained based on pixel-wise content loss functions, adversarial loss function, and a novel data-driven temporal aware loss function to maintain anatomical accuracy and temporal coherence. Besides image reconstruction, our network also performs respiratory motion compensation for free-breathing scans. A novel progressive growing-based strategy was adapted to make the training process possible for the proposed GAN-based structure. The proposed method was developed and thoroughly evaluated qualitatively and quantitatively based on 3D cardiac cine data from 42 patients. RESULTS: Our proposed method achieved significantly better scores in general image quality and image artifacts at 10.7X-15.8X acceleration than the SG CS-WV approach at 3.5X-7.9X acceleration (4.53 ± 0.540 vs. 3.13 ± 0.681 for general image quality, 4.12 ± 0.429 vs. 2.97 ± 0.434 for image artifacts, P < .05 for both). No spurious anatomical structures were observed in our images. The proposed method enabled similar cardiac-function quantification as conventional SG CS-WV. The proposed method achieved faster central processing unit-based image reconstruction (6 s/cardiac phase) than the SG CS-WV (312 s/cardiac phase). CONCLUSION: The proposed method showed promising potential for high-resolution (1 mm3 ) free-breathing 4D MR data acquisition with simultaneous respiratory motion compensation and fast reconstruction time.

CT evaluation of unrepaired/incidental congenital cardiovascular diseases in adults.

Congenital heart disease (CHD) affects approximately one million people in the USA with the number increasing by 5% each year. Patients are usually both diagnosed and treated in infancy, however many of them may have subclinical CHD that remains undiagnosed until late adulthood. Patients with complex CHD tend to be symptomatic and are diagnosed at a younger age than those with a single defect. CHDs can be divided into three categories, including cardiac, great vessels and coronary artery anomalies. Recent advances in computed tomography (CT) technology with faster acquisition time and improved spatial resolution allow for detailed evaluation of cardiac morphology and function. The concomitant increased utilization of CT has simultaneously led to more sensitive detection and more thorough diagnosis of CHD. Recognition of and understanding the imaging attributes specific to each anomaly is important for radiologists in order to make a correct and definite diagnosis. This article reviews the spectrum of CHDs, which persist into adulthood that may be encountered by radiologists on CT.

Pulmonary vein and left atrial posterior wall isolation for the treatment of atrial fibrillation: Comparable outcomes for adults with congenital heart disease.

INTRODUCTION: Optimal treatment strategies for ACHD with AF are unknown. This study sought to assess outcomes of pulmonary vein isolation (PVI) ± left atrial (LA), posterior wall isolation (PWI) for adults with congenital heart disease (ACHD), and atrial fibrillation (AF). METHODS: A retrospective review of all cryoballoon (CB) PVI ± PWI procedures at a single center over a 3-year period were performed. Clinical characteristics and outcomes for patients with and without ACHD were compared. The primary outcome was the occurrence of atrial tachyarrhythmia at 12-months postablation after a 90-day blanking period. RESULTS: Three-hundred and sixteen patients (mean: 63 ± 12 years, [63% male]) underwent CB PVI ± PWI during the study, including 31 (10%) ACHD (simple 35%, moderate 39% complex 26%; nonparoxysmal AF in 52%). ACHD was younger (51 vs. 64 years; p < .001) with a lower CHADS2 DS2 -VASc score (1.2 vs. 2.1; p = .001) but had a greater LA diameter (4.9 vs. 4.0 cm; p < .001) and a number of prior cardioversions (0.9 vs. 0.4; p < .001) versus controls. 12-month freedom from recurrent AF was similar for ACHD and controls (76% vs. 80%; p = .6) and remained nonsignificant in multivariate analysis (hazard ratio: 1.8, 95% confidence interval: 0.7-5.1; p = .22). At 12-months postablation, 75% of ACHD versus 93% of control patients were off antiarrhythmic drug therapy (p = .07). CONCLUSION: This study demonstrates younger age and lower conventional stroke risk, yet clinically advanced AF for ACHD relative to controls. CB PVI ± PWI was an effective strategy for the treatment of AF among all forms of ACHD with similar 12-month outcomes as compared to controls.

Four-dimensional Multiphase Steady-State MRI with Ferumoxytol Enhancement: Early Multicenter Feasibility in Pediatric Congenital Heart Disease.

Background The value of MRI in pediatric congenital heart disease (CHD) is well recognized; however, the requirement for expert oversight impedes its widespread use. Four-dimensional (4D) multiphase steady-state imaging with contrast enhancement (MUSIC) is a cardiovascular MRI technique that uses ferumoxytol and captures all anatomic features dynamically. Purpose To evaluate multicenter feasibility of 4D MUSIC MRI in pediatric CHD. Materials and Methods In this prospective study, participants with CHD underwent 4D MUSIC MRI at 3.0 T or 1.5 T between 2014 and 2020. From a pool of 460 total studies, an equal number of MRI studies from three sites (n = 60) was chosen for detailed analysis. With use of a five-point scale, the feasibility of 4D MUSIC was scored on the basis of artifacts, image quality, and diagnostic confidence for intracardiac and vascular connections (n = 780). Respiratory motion suppression was assessed by using the signal intensity profile. Bias between 4D MUSIC and two-dimensional (2D) cine imaging was evaluated by using Bland-Altman analysis; 4D MUSIC examination duration was compared with that of the local standard for CHD. Results A total of 206 participants with CHD underwent MRI at 3.0 T, and 254 participants underwent MRI at 1.5 T. Of the 60 MRI examinations chosen for analysis (20 per site; median participant age, 14.4 months [interquartile range, 2.3-49 months]; 33 female participants), 56 (93%) had good or excellent image quality scores across a spectrum of disease complexity (mean score ± standard deviation: 4.3 ± 0.6 for site 1, 4.9 ± 0.3 for site 2, and 4.6 ± 0.7 for site 3; P < .001). Artifact scores were inversely related to image quality (r = -0.88, P < .001) and respiratory motion suppression (P < .001, r = -0.45). Diagnostic confidence was high or definite in 730 of 780 (94%) intracardiac and vascular connections. The correlation between 4D MUSIC and 2D cine ventricular volumes and ejection fraction was high (range of r = 0.72-0.85; P < .001 for all). Compared with local standard MRI, 4D MUSIC reduced the image acquisition time (44 minutes ± 20 vs 12 minutes ± 3, respectively; P < .001). Conclusion Four-dimensional multiphase steady-state imaging with contrast enhancement MRI in pediatric congenital heart disease was feasible in a multicenter setting, shortened the examination time, and simplified the acquisition protocol, independently of disease complexity. Clinical trial registration no. NCT02752191 © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Roest and Lamb in this issue.

Case 286: Sarcoidlike Granulomatosis and Lymphadenopathy-Thoracic Manifestations of Nivolumab Drug Toxicity.

History A 70-year-old man had a posterior left thigh lesion confirmed to be biopsy-proven melanoma. The patient underwent wide excision and sentinel node biopsy, which showed absence of residual melanoma. Two years later, the patient noticed a subcentimeter subcutaneous lump in his thigh. Repeat excisional biopsy showed involvement of the surrounding soft tissue, consistent with a satellite lesion. Follow-up combined PET/CT revealed satellite nodules around the primary lesion, enabling confirmation of subcutaneous metastatic disease. The patient was subsequently started on nivolumab, an anti-programmed cell death 1 (PD-1) immune checkpoint inhibitor that blocks PD-1 and is approved as a first-line treatment in patients with advanced metastatic melanoma. On the baseline scan prior to starting nivolumab, there were no CT findings that suggested metastatic disease, nor were there enlarged mediastinal or hilar lymph nodes. Five months after initiation of nivolumab treatment, the first follow-up chest CT scan was performed and showed new findings in the mediastinum and bilateral lungs. The patient remained asymptomatic during the treatment period. Furthermore, the subcutaneous metastatic disease remained stable during the treatment period, and no other site of metastatic disease was noted on follow-up CT scans obtained during the first 5 months of treatment. The patient had no prior history of infectious or occupational exposures. During the nivolumab treatment cycle, his pertinent laboratory values and physical examination findings were unremarkable.

Case 286.

History A 70-year-old man had a posterior left thigh lesion confirmed to be biopsy-proven melanoma. The patient underwent wide excision and sentinel node biopsy, which showed absence of residual melanoma. Two years later, the patient noticed a subcentimeter subcutaneous lump in his thigh. Repeat excisional biopsy showed involvement of the surrounding soft tissue, consistent with a satellite lesion. Follow-up combined PET/CT revealed satellite nodules around the primary lesion, enabling confirmation of subcutaneous metastatic disease. The patient was subsequently started on nivolumab, an anti-programmed cell death 1 (PD-1) immune checkpoint inhibitor that blocks PD-1 and is approved as a first-line treatment in patients with advanced metastatic melanoma. On the baseline scan prior to starting nivolumab, there were no CT findings that suggested metastatic disease, nor were there enlarged mediastinal or hilar lymph nodes. Five months after initiation of nivolumab treatment, the first follow-up chest CT scan was performed and showed new findings in the mediastinum (Fig 1) and bilateral lungs (Figs 2, 3). The patient remained asymptomatic during the treatment period. Furthermore, the subcutaneous metastatic disease remained stable during the treatment period, and no other site of metastatic disease was noted on follow-up CT scans obtained during the first 5 months of treatment. The patient had no prior history of infectious or occupational exposures. During the nivolumab treatment cycle, his pertinent laboratory values and physical examination findings were unremarkable.

Chronic airspace disease: Review of the causes and key computed tomography findings.

Chronic airspace diseases are commonly encountered by chest, body or general radiologists in everyday practice. Even though there is significant overlap in the imaging findings of different causes of chronic airspace disease, some key clinical, laboratory and imaging findings can be used to guide the radiologist to the correct diagnosis. The goal of this article is to review and compare these features.

Recent Innovations in Renal Vascular Imaging.

Noninvasive imaging of the vascular renal system is a common request in diagnostic radiology. Typical indications include suspected renovascular hypertension, vasculitis, neoplasm, vascular malformation, and structural diseases of the kidney. Profound knowledge of the renal anatomy, including vascular supply and variants, is mandatory for radiologists and allows for optimized protocolling and interpretation of imaging studies. Besides renal ultrasound, computed tomography and MR imaging are commonly requested cross-sectional studies for renal and renal vascular imaging. This article discusses basic renal vascular anatomy, common imaging findings, and current and potential future imaging protocols for various renovascular pathologic conditions.

Abdominal wall and pelvic hernias: classic and unusual hernias and their mimics.

Abdominal and pelvic wall hernias are classically defined as a weakness or opening of the muscular wall through which abdominal or pelvic tissues protrude. The aim of this manuscript is to review the imaging findings of abdominal and pelvic wall hernias and their mimics and to discuss pearls and pitfalls for accurately diagnosing and classifying these entities.

Tracheobronchial Tumors: Radiologic-Pathologic Correlation of Tumors and Mimics.

Tracheobronchial masses encompass a broad spectrum of entities, ranging from benign and malignant neoplasms to infectious and inflammatory processes. This article reviews the cross-sectional findings of tracheal tumors and tumor-like entities, correlates imaging findings with histologic pathology, and discusses pearls and pitfalls in accurately diagnosing and classifying tracheal tumors and mimics.

Interatrial septum: A pictorial review of congenital and acquired pathologies and their management.

There are many different congenital abnormalities and acquired pathologies involving the interatrial septum. Differentiation of these pathologies significantly affects patient management. We have reviewed the various interatrial septal pathologies and discussed their congenital associates, clinical significance, and management. After reading this article, the reader should be able to better characterize the interatrial septal pathologies using the optimal imaging tools, and have a better understanding of their clinical significance and management.

A Private Investigation: Radiologic-Pathologic Correlation of Testicular Tumors.

To review the classification of testicular tumors, describe the sonographic and pathologic features of each tumor type, and discuss the mimics, diagnostic pitfalls, and management of testicular tumors. Method consists of pictorial review. We review sonographic and pathologic findings of several testicular tumors and tumorlike entities. Although ultrasound is the first-line imaging modality to differentiate between intratesticular and extratesticular location of an intrascrotal mass, it is not specific for intratesticular lesion characterization. Therefore, correlation with histology sampling is often necessary.