Large language models (LLMs), exemplified by generative pre-trained transformers (GPT) such as ChatGPT (OpenAI) [1], have marked a significant advancement in artificial intelligence by demonstrating exceptional capabilities in natural language processing tasks [2–4]. These models have generated considerable interest due to their potential to transform medical practice [1,5]. Recent developments have introduced multimodal-LLMs that extend beyond text analysis [6–8].
We scan the top radiology sources so you don’t have to.
From AI breakthroughs to imaging trends, we serve up real-time radiology insights.
-
Association between cognitive status and structural brain changes in Alzheimer’s disease: Clinical implication of lightweight deep learning-aided diagnosis
The complex brain changes involved in Alzheimer’s disease (AD) development constitute a high-dimensional nonlinear feature space where deep learning (DL) classification/diagnosis may be advantageous over classical non-learning methods. However, the practicality of DL remains under debate among healthcare professionals, largely because many models are computationally expensive and operate without explicit interpretability. This study aimed to construct a lightweight DL model to disclose the assoc…
-
Advancing Care for Brain Metastases in MBC
Dr Priscilla Brastianos explores the meaningful progress made in the treatment of brain metastases in metastatic breast cancer and the innovations shaping what comes next. Medscape Oncology
-
Spine Fractures to Be Flagged During Routine DEXA Scans
NICE has proposed adding vertebral fracture checks to routine DEXA scans for over-50s to detect hidden osteoporosis earlier and prevent future disability. Medscape News UK
-
Artificial intelligence-assisted CCTA for coronary stenosis detection: Diagnostic promise, methodological nuances, and directions for clinical translation
We read with great interest the recent meta-analysis by Yan et al., which comprehensively evaluates the diagnostic performance of coronary computed tomography angiography (CCTA)–based artificial intelligence (AI) systems for detecting ≥ 50 % coronary stenosis at both patient and vessel levels [1]. By rigorously adhering to PRISMA-DTA standards and disentangling patient-level from vessel-level evidence, the authors address a critical methodological limitation present in prior meta-analyses and pr…
-
Cortical vein opacification measurement using hounsfield unit values is a predictor for outcome in anterior circulation acute ischemic stroke
In acute ischemic stroke (AIS) patients, the cortical vein opacification on computed tomographic angiography (CTA) imaging is usually asymmetric. To explore the correlation between CT perfusion parameters and cortical vein opacification evaluated by Hounsfield unit (HU) values, which is more accurately, and compare the predictive ability between them for outcome in AIS.
-
VIPs outperforms established models for predicting post-TIPS prognosis in viral hepatitis-dominant cirrhosis
Transjugular intrahepatic portosystemic shunt (TIPS) manages portal hypertension complications in cirrhosis, but predicting post-TIPS outcomes remains challenging, especially in viral hepatitis-dominated populations.
-
Nerve Ablation for Knee OA: Where Is Its Place in Treatment?
In an expert Q&A, Hospital for Special Surgery’s Lisa Mandl, MD, MPH, discusses the role of genicular nerve ablation in reducing knee OA symptoms amidst other nonpharmacologic options. Hospital for Special Surgery