AI Blog

Impact of deep learning reconstruction on radiation dose reduction and cancer risk in CT examinations

Takeshi Nakaura April 2025

Deep‑learning reconstruction (DLR) shifts CT image formation from a hardware‑limited process to a data‑driven one. In our real‑world cohort of >10,000 body scans, we observed a 45% fall in effective dose and a halving of lifetime attributable cancer risk without sacrificing diagnostic confidence. The message is simple: When the image quality can be algorithmically recovered, radiologists are no longer forced to

ChatGPT-4o powers AI-generated simplified breast imaging reports for enhanced patient comprehension

Roberto Maroncelli April 2025

In our study, we evaluated ChatGPT-4o’s ability to simplify breast imaging reports while maintaining accuracy and completeness. Radiology reports often contain complex terminology, creating barriers to patient understanding. We aimed to assess whether AI-driven simplifications could enhance clarity without introducing errors. Our results show that AI-generated summaries maintained high factual accuracy (median score: 2/5) and completeness (2/5), with a low

How to get to valuable radiology AI: the role of early health technology assessment

Erik Kemper April 2025

Artificial intelligence (AI) technologies, such as large language and deep learning models, are poised to significantly influence clinical practice. However, the real-world impact and utility of these AI applications, particularly in radiology, often remain unclear. How can their impact be measured, and can we ensure it benefits everyone? In ESR Essentials: how to get to valuable radiology AI: the role

Impact of uncertainty quantification through conformal prediction on volume assessment from DL-based MRI prostate segmentation

Insights into Imaging March 2025

This study aimed to evaluate the uncertainty of a deep learning (DL)-based prostate segmentation algorithm using conformal prediction (CP) and its impact on prostate volume (PV) calculation in patients at risk of prostate cancer (PC). The study involved 377 patients’ 3-Tesla T2-weighted scans. By applying CP at an 85% confidence level, unreliable pixel segmentations of the DL model were flagged,

This blur detection model could provide instantaneous feedback to technicians

European Radiology Experimental March 2025

This study developed a model to automatically detect blurred areas in mammograms, which can affect diagnostic accuracy. Using a retrospective dataset consisting of 152 mammograms from three vendors, expert radiologists outlined blurred regions. Normalized Wiener spectra (nWS) were extracted and processed through a convolutional neural network (CNN) to classify images as either blurred or sharp. The model showed an AUROC

On Artificial Intelligence: An interview with Gennaro D’Anna

Alex Vogel, Gennaro D'Anna March 2025

The latest addition to our “On Artificial Intelligence” series finds us diving into artificial intelligence (AI) within the world of radiology and, more specifically, neuroradiology. Gennaro D’Anna, a neuroradiologist at CDI Centro Diagnostico Italiano in Milan, Italy, discusses challenges to adoption in clinical practice, what the future looks like for AI, and what new generations of radiologists can do to

AI and machine learning-based diagnostic advantages and pitfalls: Musculoskeletal fracture detection

European Radiology March 2025

Artificial intelligence (AI) is increasingly integrated into healthcare, especially within diagnostic radiology. AI algorithms, particularly those utilizing machine learning and computer vision, are widely used for evaluating MRI, CT, and radiographic images. These models, trained on extensive patient databases, are now fully commercialized. Ongoing research compares AI’s efficacy to human radiologists in diagnosing various pathologies, such as scoliosis, hip dysplasia,

Avoiding Radiology’s Boeing 737 Max moment

Raymond Geis March 2025

Recently, Kemper et al. published an insightful paper on the challenges and opportunities of health technology assessment (HTA) models to evaluate the value, safety, and trustworthiness of radiology computer vision AI (RCVAI) tools used in daily practice. Formal HTA frameworks are theoretically accurate and help define complex issues, but they are ideal approaches. The real world is messier. Ideal descriptions

Image biomarkers and explainable AI

European Radiology Experimental February 2025

Feature extraction and selection in medical data are crucial for radiomics and image biomarker discovery, particularly using convolutional neural networks (CNNs). The process involves feature extraction, dimensionality reduction, and addressing the curse of dimensionality. While deep learning (DL) techniques perform well, handcrafted features are important for certain studies and need to be considered. Dataset size and diversity are also key

On Artificial Intelligence: An interview with Daniel Truhn

Alex Vogel, Daniel Truhn February 2025

Following a fellowship in artificial intelligence (AI) at the Institute of Imaging and Computer Vision at RWTH Aachen, which only solidified his belief in AI’s great potential in medicine, Daniel Truhn, a Diagnostic and Interventional radiologist at University Hospital Aachen in Germany, started working on AI’s application in medicine, even creating his own working group in the area. Join us

Welcome to the blog on Artificial Intelligence of
the European Society of Radiology

Latest posts