Radiology最新文献

The Mammography Quality Standards Act (MQSA) of 1992 is intended to ensure that mammography practice nationwide meets consistent baseline quality standards. Amendments to the MQSA implementing regulations ("Amendments") were published on March 10, 2023, and are effective on September 10, 2024. The Amendments address various aspects of the program, including mammography technology, enforcement, the retention and transfer of personnel records and medical records, the medical outcomes audit, and mammography reporting, including (but not limited to) reporting of breast tissue density. The amended regulations are available online, and the Food and Drug Admininstration (FDA) offers several resources for mammography facilities and other stakeholders to receive additional information, including a facility hotline, a summary document distributed to all certified mammography facilities, and a Small Entity Compliance Guide (or SECG) written in question-and-answer format, which the FDA intends to be helpful to facilities of any size.

Background MRI plays a crucial role in restaging locally advanced rectal cancer treated with total neoadjuvant therapy (TNT); however, prospective studies have not evaluated its ability to accurately select patients for nonoperative management. Purpose To evaluate the ability of restaging MRI to predict oncologic outcomes and identify imaging features associated with residual disease (RD) after TNT. Materials and Methods This was a secondary analysis of the Organ Preservation in Rectal Adenocarcinoma (OPRA) trial, which randomized participants from April 2014 to March 2020 with stages II or III rectal adenocarcinoma to undergo either induction or consolidation TNT. Participants enrolled in the OPRA trial who underwent restaging MRI were eligible for inclusion in the present study. Radiologists classified participants as having clinical complete response (cCR), near-complete clinical response (nCR), or incomplete clinical response (iCR) based on restaging MRI at a mean of 8 weeks ± 4 (SD) after treatment. Oncologic outcomes according to MRI response category were assessed using Kaplan-Meier curves. Logistic regression analysis was performed to identify imaging characteristics associated with RD. Results A total of 277 participants (median age, 58 years [IQR, 17 years]; 179 male) who were randomized in the OPRA trial had restaging MRI forms completed. The median follow-up duration was 4.1 years. Participants with cCR had higher rates of organ preservation compared with those with nCR (65.3% vs 41.6%, log-rank P < .001). Five-year disease-free survival for participants with cCR, nCR, and iCR was 81.8%, 67.6%, and 49.6%, respectively (log-rank P < .001). The MRI response category also predicted overall survival (log-rank P < .001), distant recurrence-free survival (log-rank P = .005), and local regrowth (log-rank P = .02). Among the 266 participants with at least 2 years of follow-up, 129 (48.5%) had RD. At multivariable analysis, the presence of restricted diffusion (odds ratio, 2.50; 95% CI: 1.22, 5.24) and abnormal nodal morphologic features (odds ratio, 5.04; 95% CI: 1.43, 23.9) remained independently associated with RD. Conclusion The MRI response category was predictive of organ preservation and survival. Restricted diffusion and abnormal nodal morphologic features on restaging MRI scans were associated with increased likelihood of residual tumor. ClinicalTrials.gov identifier: NCT02008656 © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Milot in this issue.

Background Recent advancements, including image processing capabilities, present new potential applications of large language models such as ChatGPT (OpenAI), a generative pretrained transformer, in radiology. However, baseline performance of ChatGPT in radiology-related tasks is understudied. Purpose To evaluate the performance of GPT-4 with vision (GPT-4V) on radiology in-training examination questions, including those with images, to gauge the model's baseline knowledge in radiology. Materials and Methods In this prospective study, conducted between September 2023 and March 2024, the September 2023 release of GPT-4V was assessed using 386 retired questions (189 image-based and 197 text-only questions) from the American College of Radiology Diagnostic Radiology In-Training Examinations. Nine question pairs were identified as duplicates; only the first instance of each duplicate was considered in ChatGPT's assessment. A subanalysis assessed the impact of different zero-shot prompts on performance. Statistical analysis included χ² tests of independence to ascertain whether the performance of GPT-4V varied between question types or subspecialty. The McNemar test was used to evaluate performance differences between the prompts, with Benjamin-Hochberg adjustment of the P values conducted to control the false discovery rate (FDR). A P value threshold of less than.05 denoted statistical significance. Results GPT-4V correctly answered 246 (65.3%) of the 377 unique questions, with significantly higher accuracy on text-only questions (81.5%, 159 of 195) than on image-based questions (47.8%, 87 of 182) (χ² test, P < .001). Subanalysis revealed differences between prompts on text-based questions, where chain-of-thought prompting outperformed long instruction by 6.1% (McNemar, P = .02; FDR = 0.063), basic prompting by 6.8% (P = .009, FDR = 0.044), and the original prompting style by 8.9% (P = .001, FDR = 0.014). No differences were observed between prompts on image-based questions with P values of .27 to >.99. Conclusion While GPT-4V demonstrated a level of competence in text-based questions, it showed deficits interpreting radiologic images. © RSNA, 2024 See also the editorial by Deng in this issue.

Background Podcasts have become an increasingly popular method of communicating information in medicine, including in radiology. However, the effect of podcasts on the reach of journal articles remains unclear. Purpose To evaluate the influence of Radiology podcasts on the performance metrics, including downloads, citations, and Altmetric Attention Score (AAS), of Radiology articles. Materials and Methods This was a retrospective study. All articles published in the print version of Radiology from January 2021 to December 2022 were reviewed; editorials and case reports were excluded. Articles featured on Radiology podcasts were included in the podcast group. Articles published within the same journal issue and category were the nonpodcast group. Downloads, Google Scholar citations, Dimensions citations, and AAS metrics were recorded. The Mann-Whitney U test was used to compare medians and evaluate differences between older and more recently published articles. Results The podcast group, composed of 88 articles, exhibited significantly higher median values for downloads (PG, 4521.0; nonpodcast group, 2123.0; P < .001), Google Scholar citations (podcast group, 14.5; nonpodcast group, 10.0; P = .01), Dimensions citations (podcast group, 12.0; nonpodcast group, 9.0; P = .01), and AAS (podcast group, 43.0; nonpodcast group, 10.0; P < .001) compared with the nonpodcast group of 378 articles. Within both groups, articles published in the earlier nonpodcast group (January to June 2021) had higher downloads (podcast group, P = .08; nonpodcast group, P < .001), Google Scholar citations (podcast group and nonpodcast group, P < .001), and Dimension citations (podcast group and nonpodcast group, P < .001) than articles from the later period (July to December 2022). AAS markedly increased in recent podcast articles (P = .03), but AAS for nonpodcast articles significantly decreased over time (P = .01). Conclusion Radiology articles featured on the Radiology podcast had greater median metrics, including downloads, Google Scholar citations, Dimensions citations, and AAS, compared with nonpodcast articles, suggesting that podcasts can be an effective method of disseminating and amplifying research within the field of radiology. © RSNA, 2024 See also the editorial by Chu and Nicola in this issue.