AI Blog

Prediction of lipomatous soft tissue malignancy on MRI

European Radiology Experimental February 2023

This study, which included a cohort of 145 patients affected by lipomatous soft tissue tumours, aimed to compare the performances of MRI radiomic machine learning analysis with that of deep learning in order to predict malignancy in patients with lipomas oratypical lipomatous tumours. The authors were able to show that batch-effect corrected machine learning and radiomics approaches outperformed deep learning-based

Opportunistic screening for osteoporosis and osteopenia using machine learning

Ronnie Sebro January 2023

Dr. Rizwan Aslam, of the University of California, San Francisco (UCSF), presented an abstract at RSNA in 2008 which showed that it was possible to screen for osteoporosis from CT colonoscopy scans while I was a resident at UCSF. Subsequently, other papers showed that there were strong correlations between CT measurements and the measurements obtained from dual-energy x-ray absorptiometry (DXA).

Application of deep learning reconstruction in the diagnosis of renal calculi

Insights into Imaging January 2023

In this study, the authors evaluated the diagnostic capability, image quality, and radiation dose of abdominal ultra-low-dose CT (ULDCT) with deep learning reconstruction (DLR) in the detection of renal calculi, which are a common urological disease and normally detected by CT. They were able to find that ULDCT combined with DLR could significantly reduce radiation dose, decrease image noise, and

Using parametric feature maps to enhance the stability of CT radiomics

Laura Jensen, Damon Kim, Sebastian Nagel January 2023

It is known that the reproducibility of radiomic features is influenced by myriad factors, one of which is the size of the segmented volume. We hypothesized that parametric maps calculated with a fixed voxel size could address this issue. To test the hypothesis, we conducted a phantom study and could show that the stability across different volumes of interest sizes

Cancer-associated incidental pulmonary embolism – underreporting and the potential role of AI

Peder Wiklund January 2023

Pulmonary embolism (PE) is a common complication in patients with cancer. A significant number of all PE are diagnosed incidentally (incidental PE, iPE) in CT scans performed for staging or treatment response evaluation. Thousands of CT scans are performed for this indication every year in Halmstad, a regional hospital in Region Halland in southern Sweden, and we anticipate that in

Deep learning image reconstruction to improve accuracy of iodine quantification and image quality in abdomen dual-energy CT

European Radiology January 2023

The aim of this study was to investigate the effect of deep learning image reconstruction (DLIR) on the accuracy of iodine quantification and image quality of dual-energy CT (DECT) compared to other reconstruction algorithms. The authors showed that the DLIR algorithm reduced image noise and variability of iodine concentration values when compared with other reconstruction algorithms. Key points In the

Reproducibility of AI models in head CT

Insights into Imaging December 2022

The authors of this study reviewed research on AI algorithms relating to computed tomography (CT) of the head in order to verify to what degree it is true that AI software for applications in radiology must be transferable to other real-world problems. It was discovered that current research on AI for head CT is rarely reproducible, does not match with

The role of radiomics in the detection of lymph node metastases

European Radiology Experimental December 2022

The authors of this retrospective analysis looked at the role that radiomics played when applied to contrast-enhanced computed tomography (CT) in detecting lymph node (LN) metastases in lung cancer patients compared to 18F-fluorodeoxyglucose positron emission tomography (PET)/CT as a reference. The authors determined that radiomics showed good discrimination power, regardless of the modelling technique, in detecting LN metastases in lung

Machine learning–based radiomics classifies parotid tumors using morphological MRI

European Radiology December 2022

This comparative study aimed to evaluate the effectiveness of machine learning models based on morphological MRI radiomics in the classification of parotid tumors. The authors developed three-step machine learning models with extreme gradient boosting (XGBoost), support vector machine (SVM), and decision tree (DT) algorithms in order to classify the parotid neoplasms into four subtypes. The study was able to demonstrate

Communicating with patients in the age of online portals

Christoph Becker December 2022

Future generations of patients are very likely to get more involved in decisions regarding their healthcare according to the motto “nothing about me without me”. The deployment of electronic patient portals increasingly allows patients throughout Europe to consult and share their medical data, including radiology reports and images, securely and timely online. Technical solutions and rules for releasing reports and

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

Latest posts