AI Blog

AI for prostate MRI: open datasets, available applications, and grand challenges

European Radiology Experimental April 2023

This narrative review provides an overview of the current state-of-the-art artificial intelligence (AI) applications for prostate MRI by focusing on open datasets, commercially and publically available AI systems, and challenges. The authors state that large amounts of research are still required in order to successfully utilize AI in the whole prostate pathway. Due to the rapidly growing field, continuous up-to-date

Quality assurance for automatically generated contours with additional deep learning

Insights into Imaging April 2023

This study explores the importance of quality assurance when deploying an automatic segmentation model. The authors of this study built a deep-learning model with the goal of estimating the quality of automatically generated contours. They found that the trained model can be used alongside automatic segmentation tools, thus ensuring quality and allowing intervention to prevent undesired segmentation behavior. Key points

Application of deep learning to improve image quality and reduce scan time

European Radiology April 2023

In this study, the image quality and diagnostic performance of conventional motion-corrected periodically rotated overlapping parallel line with enhanced reconstruction (PROPELLER) MRI sequences was compared with post-processed PROPELLER MRI sequences using deep learning (DL)-based reconstructions. The authors found that the accelerated PROPELLER sequences with DL post-processing showed superior image quality and higher diagnostic confidence when compared to conventional PROPELLER sequences.

Tasks for AI in prostate MRI

European Radiology Experimental March 2023

The authors of this narrative review aimed to introduce quality metrics for emerging artificial intelligence (AI) papers, such as the Checklist for Artificial Intelligence in Medical Imaging (CLAIM) and Field-Weighted Citation Impact (FWCI). Furthermore, the study dives into some of the top AI models for segmentation, detection, and classification, while concluding that prospective studies with multi-center design will need to

Utilizing CLAIM to adapt the increasing trend of deep learning application in radiomics

Insights into Imaging March 2023

The aim of this study was to provide an updated systematic review of radiomics in osteosarcoma, utilizing various databases such as PubMed, Embase, China National Knowledge Infrastructure, and more. Articles found in these databases were assessed by Radiomics Quality Score (RQS), Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement, Checklist for Artificial Intelligence in

Detecting dysthyroid optic neropathy using radiomics analysis of the optic nerve

Insights into Imaging March 2023

Because the early detection of dysthyroid optic neuropathy (DON) is of the utmost importance for clinical decision-making, the authors of this study aimed to determine the feasibility of using an optic-nerve-based radiomics nomogram on water-fat imaging for its detection. It was found that the optic-nerve-based radiomics nomogram displayed better diagnostic performance when compared to conventional MRI evaluation. Key points Radiomics

Assessment of automatic rib fracture detection on chest CT using a deep learning algorithm

European Radiology March 2023

This retrospective study evaluated deep learning algorithms for the detection of automatic rib fracture on thoracic CT scans. The authors also aimed to compare its performance with attending-level radiologists using an internal dataset of 12,208 ER trauma patients and an external dataset of 1,613 ER trauma patients taking chest CT scans. The study showed that the proposed deep learning model

The effect of preprocessing filters on predictive performance in radiomics

Aydin Demircioğlu February 2023

In radiomics, the main goal is to extract quantitative features from medical data to train a predictive model using machine learning techniques. Contrary to statistics, radiomics is data-driven; thus, there is a certain tendency to use as many features as possible. To achieve this, preprocessing filters are often applied; for example, a Gaussian filter smooths the image and thus removes

EuSoMII Radiomics Auditing Group Initiative: Review of RQS applications

Arnaldo Stanzione February 2023

When the Radiomics Quality Score (RQS) was presented to the scientific community back in 2017, its authors aimed to introduce a tool for a rapid and effective evaluation of radiomics studies’ scientific/clinical merit. Conceived as a quality seal to be published alongside presented results and newly proposed radiomics models, the RQS has instead mostly been adopted by researchers as a

Standardised lesion segmentation for imaging biomarker quantitation

Nandita deSouza February 2023

Extraction of quantitative data from images requires selecting the regions-of-interest from which this data is to be extracted. This process of region-of-interest selection (or segmentation), whether it be manual or automated, crucially determines the values of the extracted biomarkers. Harmonisation of the segmentation process itself is therefore central to the standardisation of imaging biomarkers. We undertook a Delphi process amongst

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

Latest posts