AI Blog

Clinical applications of AI and radiomics in neuro-oncology

Insights into Imaging May 2022

In this educational review, the authors take a comprehensive look at various aspects and applications of artificial intelligence (AI) in the field of neuro-oncology, including machine learning, deep learning, and radiomics. The merits and challenges of the deployment and use of AI tools in neuro-oncology are put under the microscope, with the authors concluding that AI has a promising future

Can machine learning predict the WHO/ISUP nuclear grade of clear cell renal cell carcinomas?

Insights into Imaging May 2022

In this retrospective study, the authors investigated if radiomics features extracted from nephrographic-phase (NP) CT images combined with clinicoradiological characteristics may have the potential to preoperatively differentiate between low- and high-nuclear grade of clear cell renal cell carcinomas (CCRCCs). They were able to demonstrate that radiomics analysis may be used as a potentially noninvasive method for distinguishing low- from high-grade

Radiomics in the prediction of disease-free survival in early-stage squamous cervical cancer

European Radiology May 2022

The authors of this study conducted multiparametric magnetic resonance imaging (MRI)-derived radiomics based on multi-scale tumor region in order to predict disease-free survival (DFS) in a cohort of 191 patients with early-stage squamous cervical cancer (ESSCC). They were able to conclude that multiparametric MRI-derived radiomics based on multi-scale tumor region can in fact aid in the prediction of DFS for

Creating a training set for AI from initial segmentations of airways

Ivan Dudurych April 2022

An important challenge in the use of artificial intelligence (AI) for medical image segmentation tasks is the lack of high-quality, scan protocol-specific datasets. AI performs best on narrow tasks with homogenous specifications. Thus, pre-trained models may be inadequate for use in centre-specific studies if the scan protocols do not match. For the airway segmentation task in the Imaging in Lifelines

A deep learning algorithm for VHD diagnosis and evaluation

European Radiology April 2022

This study aims to develop and validate a deep learning-based automatic chest radiograph (CXR) cardiovascular border (CB) analysis algorithm (CB_auto) in order to diagnose and quantitatively evaluate valvular heart disease (VHD). The authors found that the CB_auto system, in coordination with the deep learning algorithm, provided highly reliable CB measurements, which, in turn, can be useful, not long daily clinical

Deep learning radiomics in prediction of NAC response

European Radiology April 2022

For breast cancer, the standard of treatment for most patients is neoadjuvant chemotherapy (NAC), but response rates may vary among patients, causing delays in appropriate treatment. The authors of this prospective study aimed to investigate the feasibility of deep learning radiomics (DLR) in the prediction of NAC response at an early stage of breast cancer treatment. The authors found that

Misunderstandings on equitable and inequitable biases in machine learning

Insights into Imaging April 2022

In this reply to a ‘Letter to the editor‘, the authors seek to correct a few misunderstandings relating to equitable and inequitable biases in machine learning and radiology that were mentioned by the authors of the letter, in which topics such as cultural biases and ‘socially related’ cognitive biases were discussed, as well as how to deal with these biases.

Deep learning assesses additional radiation dose in overscanning

Yazdan Salimi March 2022

Following the COVID-19 pandemic, the number of chest CT examinations has dramatically increased, which will undeniably impact public medical exposure. Overscanning, i.e., scanning unnecessary regions in the axial field-of-view, causes noticeable excessive radiation dose to patients undergoing chest CT examinations. The manual procedure of selecting the scan range based on anterior-posterior or lateral localizers is prone to human error in

Can AI predict breast tumour response?

European Radiology March 2022

This proof of concept study examines using a deep learning-based method for the automatic analysis of digital mammograms as a tool to aid in the assessment of neoadjuvant chemotherapy (NACT) treatment response to breast cancer. The authors found that the initial AI performance was able to indicate the potential to aid in clinical decision-making, but in order to continue exploring

Radiomic nomogram predicts pathology invasiveness

European Radiology March 2022

The purpose of this retrospective diagnostic study was to develop and validate a preoperative CT-based nomogram combined with radiomic and clinical-radiological signatures to distinguish preinvasive lesions from pulmonary invasive lesions. The authors were able to demonstrate that a nomogram constructed by identified clinical-radiological signatures and combined radiomic signatures has the potential to precisely predict pathology invasiveness. Key points The radiomic

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

Latest posts