Acad Radiol
MRI unit, Radiology department. HT medica. Carmelo Torres nº2, 23007, Jaén, Spain (T.M.N., A.L.).
Published: December 2024
Rationale And Objectives: Large Language Models can capture the context of radiological reports, offering high accuracy in detecting unexpected findings. We aim to fine-tune a Robustly Optimized BERT Pretraining Approach (RoBERTa) model for the automatic detection of unexpected findings in radiology reports to assist radiologists in this relevant task. Second, we compared the performance of RoBERTa with classical convolutional neural network (CNN) and with GPT4 for this goal.
Materials And Methods: For this study, a dataset consisting of 44,631 radiological reports for training and 5293 for the initial test set was used. A smaller subset comprising 100 reports was utilized for the comparative test set. The complete dataset was obtained from our institution's Radiology Information System, including reports from various dates, examinations, genders, ages, etc. For the study's methodology, we evaluated two Large Language Models, specifically performing fine-tuning on RoBERTa and developing a prompt for ChatGPT. Furthermore, extending previous studies, we included a CNN in our comparison.
Results: The results indicate an accuracy of 86.15% in the initial test set using the RoBERTa model. Regarding the comparative test set, RoBERTa achieves an accuracy of 79%, ChatGPT 64%, and the CNN 49%. Notably, RoBERTa outperforms the other systems by 30% and 15%, respectively.
Conclusion: Fine-tuned RoBERTa model can accurately detect unexpected findings in radiology reports outperforming the capability of CNN and ChatGPT for this task.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1016/j.acra.2024.07.057 | DOI Listing |
Clin Imaging
December 2024
Department of Radiology, Hospital Sirio-Libanes, Rua Adma Jafet, 91, Sao Paulo, SP 01308-050, Brazil.
Radiofrequency ablation is a treatment option for primary and secondary lung tumors, particularly in non-surgical patients. Although it is considered a safe and effective procedure, various complications can occur. Most of these adverse events are self-limiting and can be treated conservatively; however, some may be severe and require intervention.
View Article and Find Full Text PDFJ Clin Med
January 2025
Faculty of Medicine, "Carol-Davila" University of Medicine and Pharmacy, 050474 Bucharest, Romania.
: Bupropion, an atypical antidepressant and smoking cessation aid, is known for its potential to cause seizures, cardiotoxicity and neurotoxicity in overdose scenarios. However, overdoses may present variably, and muscular and renal complications, such as rhabdomyolysis and acute kidney injury (AKI), can emerge in unexpected ways. Previous reports have shown that severe overdoses can lead to a spectrum of complications, but the precise mechanisms linking bupropion overdose with rhabdomyolysis remain poorly understood.
View Article and Find Full Text PDFMolecules
December 2024
Department of Physical Pharmacy and Pharmacokinetics, Poznan University of Medical Sciences, 3 Rokietnicka Street, 60-806 Poznań, Poland.
Tryptophan (TRP) is an essential amino acid crucial for the production of many bioactive compounds. Disturbances in TRP metabolism have been revealed in various diseases, many of which are closely related to the immune system. In recent years, we have focused on finding blood-based biomarkers of successful immunotherapy in cancer.
View Article and Find Full Text PDFNat Microbiol
January 2025
Program in Emerging Infectious Diseases, Duke-NUS Medical School, Singapore, Singapore.
T cells have been identified as correlates of protection in viral infections. However, the level of vaccine-induced T cells needed and the extent to which they alone can control acute viral infection in humans remain uncertain. Here we conducted a double-blind, randomized controlled trial involving vaccination and challenge in 33 adult human volunteers, using the live-attenuated yellow fever (YF17D) and chimeric Japanese encephalitis-YF17D (JE/YF17D) vaccines.
View Article and Find Full Text PDFLupus Sci Med
January 2025
Department of Otolaryngology-Head and Neck Surgery, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China
Objective: The present study aimed to provide a comprehensive evaluation of the postmarketing safety of belimumab based on the Food and Drug Administration Adverse Event Reporting System (FAERS) database.
Methods: Adverse event (AE) reports in the FAERS database from January 2021 to December 2023 were extracted to perform the disproportionality analysis by calculating the reporting OR. The clinical characteristics and onset times of AEs were investigated.
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!
© LitMetric 2025. All rights reserved.