A CT-based AI model identified individuals with stage IA non–small cell lung cancer who underwent segmentectomy and were at high risk for recurrence and low overall survival. SNU Radiology bit.ly/3xwwFqk

Using data manually extracted by a radiologist as the ground truth, GPT-4 more frequently extracted correct data from free-text reports on mechanical thrombectomy in patients with ischemic stroke than GPT-3.5. Daniel Paech bit.ly/4aVsumn

Large language model GPT-4 matched the performance of radiologists in detecting errors in radiology reports, according to research published in Radiology. bit.ly/3Q8QJpd

CT images before interventional therapy show reduced arterial liver perfusion, decreased portal venous perfusion, and elevated hepatic perfusion index. DSA confirmed stenosis of left hepatic vein and successful restoration of vessel patency. bit.ly/4aCi6QD

In this editorial, Drs. Jae Ho Sohn, MD, MS and Brandon K.K. Fields, M.D. discuss the study by Pan and Hu et al for predicting preinvasive, minimally invasive, and invasive adenocarcinoma. UCSF Radiology Residents UCSF Imaging UCSF Center for Intelligent Imaging @UCSF @UCSFchest doi.org/10.1148/radiol…

Performance of DL models for predicting preinvasive, minimally invasive, or invasive lung adenocarcinoma was improved by combining binary and ternary classification models for predicting invasiveness and adjudication of discordant classification. bit.ly/49oR6Tq

🚨 A new episode of the Radiology podcast is now available!🚨 Linda Chu, MD speaks scientific editor Sarah Atzen about tips for writing materials and methods in Radiology #RadPodcasts bit.ly/3QZQRsC

The SRU expert panel recommends performing an augmented pelvic US examination in premenopausal or early postmenopausal individuals (<5 years since the cessation of menses) who are symptomatic for or have a history of endometriosis or infertility.​ bit.ly/49zXFCG

In this targeted editorial, Drs. KimMyoungKyoung and Jung Min Chang discuss the selection of patients who could benefit from supplemental imaging for breast cancer and various tools, including AI models, that can assist in this clinical task. doi.org/10.1148/radiol…

AISmartDensity, an AI-based score combining cancer signs, masking, and risk AI models, outperformed density models in selecting patients who could benefit from supplemental imaging for breast cancer. bit.ly/3VMQpQI

MRI scan shows a mass with high signal intensity involving the right orbit, middle cranial fossa, and subcutaneous tissue, while carotid DSA demonstrates heterogeneous enhancement of the lesion. bit.ly/3PXmVM9

Compared with guidelines, supersonic shear imaging liver stiffness dual threshold values of <7 kPa and >12 kPa were close to target sensitivity and specificity for compensated advanced chronic liver disease in high-risk patients. pubs.rsna.org/doi/10.1148/ra…

Compared with guidelines, supersonic shear imaging liver stiffness dual threshold values of <7 kPa and >12 kPa were close to target sensitivity and specificity for compensated advanced chronic liver disease in high-risk patients. bit.ly/3TPOiJh

📢 Breast Cancer Screening
Are you curious about how AI can assist in identifying patients who may benefit from supplemental imaging after negative mammography?

Read more about #AISmartDensity at doi.org/10.1148/radiol….

Radiology Rad In Training Editor Victoria Chernyak, MD MS FSAR FESGAR FACR
#RadInTraining

Eastern HCC diagnostic guidelines offered higher sensitivity, whereas Western guidelines provided higher specificity in patients with cirrhosis and in those with chronic hepatitis B. bit.ly/4avuX6W

🚨 A new episode of the Radiology podcast is now available! 🚨 Linda Chu, MD and Dr. Kim speak with Cooky Menias and Amy Hara about Women Leaders in Radiology #RadPodcasts bit.ly/3QZQRsC

Dr. Jonathan Mezrich of Yale Department of Radiology & Biomedical Imaging discusses this retrospective analysis of 135 teleradiology medical malpractice cases and 3474 radiology medical malpractice cases not involving teleradiology. doi.org/10.1148/radiol…

Teleradiology medical malpractice cases more frequently involved the death of the patient and had higher median indemnity payment amounts compared with cases not involving teleradiology. Aaron Sodickson MD, PhD, FASER bit.ly/4aDUElJ