AI Blog

COVID-19 classification of X-ray images using deep neural networks

European Radiology September 2021

The authors of this retrospective study propose a deep learning model for the detection of COVID-19 from chest x-rays (CXRs), as well as a tool for retrieving similar patients according to the model’s results on their CXRs. The data used for training and evaluating this model was collected from inpatients across four different hospitals. The proposed model achieved accuracy of

AI-based improvement in lung cancer detection on chest radiographs: results of a multi-reader study in NLST dataset

Hyunsuk Yoo, Mannudeep Kalra, Subba Digumarthy September 2021

Our study included 519 screening chest radiographs (CXRs) from 294 patients enrolled in the National Lung Screening Trial (NLST) who either had proven to have lung cancer or did not have lung cancer over the duration of the trial. Five attending radiologists and three radiology residents from South Korea and the U.S. independently assessed all CXRs for the presence of

Machine learning based on the multimodal connectome can predict the preclinical stage of Alzheimer’s disease: a preliminary study

European Radiology September 2021

Subjective cognitive decline (SCD) may be a preclinical stage of Alzheimer’s disease (AD). For this preliminary study, the authors recruited one hundred ten SCD individuals and well-matched healthy controls (HCs) in order to find if machine learning based on the multimodal connectome could predict the preclinical stage of AD. The study found that the characteristics identified from the multimodal network

Fully automated pelvic bone segmentation in multiparametric MRI using a 3D convolutional neural network

Xiaoying Wang August 2021

The accurate skeleton segmentation with their semantic labels represents the initial step to achieve accurate prostate cancer bone metastases detection on DWI and ADC images. Several studies have reported convolutional neural networks (CNNs) for the segmentation of normal bone structures on CT images and bone scans; however, only a few studies on automatic segmentation of normal bone structures on MR

Practice makes perfect: Using AR to simulate CT guided procedures

Dimitri Amiras August 2021

As a junior radiology trainee, when I was not in the reporting room or on the wards doing portable ultrasounds, I found learning CT procedures a daunting task for the first time on a patient. This Halsted method of training, whilst effective, is expensive, takes a long time to achieve competency and can be a stressful experience. Simulation training, for

Automated segmentation and quantification of the healthy and diseased aorta in CT angiographies using a dedicated deep learning approach

European Radiology August 2021

Performed between 2015 and 2018, the purpose of this study was to develop and validate a deep learning-based algorithm for the segmentation and quantification of the physiological and diseased aorta in computed tomography (CT) angiographies. The authors were able to determine that automated segmentation of the aorta on CTA data using a deep learning algorithm is feasible and accurate, relating

Artificial intelligence for prediction of COVID-19 progression using CT imaging and clinical data

European Radiology August 2021

Although challenging to predict, early recognition of COVID-19 severity can help guide patient management. The authors of this study aimed to develop an artificial intelligence system that was capable of predicting future deterioration to critical illness in COVID-19 patients. The AI system was developed to integrate chest CT and clinical data for risk prediction of said future deterioration to critical

Deep learning–based algorithm to detect primary hepatic malignancy in multiphase CT of patients at high risk for HCC

So Yeon Kim July 2021

Of late, deep learning-based algorithms have been successfully applied to various medical imaging modalities, ranging from chest radiographs to head CT scans. Compared to other body parts, there is a paucity of data regarding the application of deep learning-based algorithms in the liver. This can be attributed to the following reasons: First, unlike other body parts usually relying on single-phase

Combined radiomics-clinical model to predict malignancy of vertebral compression fractures on CT

European Radiology July 2021

The authors of this study aimed to develop and validate a combined radiomics-clinical model to predict malignancy of cerebral compression fractures on CT. This study comprised 165 patients with vertebral compression fractures who were allocated to training and validation cohorts. The authors were able to determine that the combined radiomics-clinical model integrating clinical parameters with radiomics score could predict malignancy

A comparison between manual and artificial intelligence–based automatic positioning in CT imaging for COVID-19 patients

European Radiology July 2021

The authors of this study analyzed and compared the imaging workflow, radiation dose, and image quality for 127 adult COVID-19 patients who were examined using either the conventional manual positioning (MP) method or an AI-based automatic positioning (AP) method. The patients underwent chest CT scans using the manual positioning (MP group) for the initial scan, followed by an AI-based automatic

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

Latest posts