Artificial Intelligence in Diagnostic Radiology: Where Do We Stand, Challenges, and Opportunities

Abstract

Artificial intelligence (AI) is the most revolutionizing development in the health care industry in the current decade, with diagnostic imaging having the greatest share in such development. Machine learning and deep learning (DL) are subclasses of AI that show breakthrough performance in image analysis. They have become the state of the art in the field of image classification and recognition. Machine learning deals with the extraction of the important characteristic features from images, whereas DL uses neural networks to solve such problems with better performance. In this review, we discuss the current applications of machine learning and DL in the field of diagnostic radiology.Deep learning applications can be divided into medical imaging analysis and applications beyond analysis. In the field of medical imaging analysis, deep convolutional neural networks are used for image classification, lesion detection, and segmentation. Also used are recurrent neural networks when extracting information from electronic medical records and to augment the use of convolutional neural networks in the field of image classification. Generative adversarial networks have been explicitly used in generating high-resolution computed tomography and magnetic resonance images and to map computed tomography images from the corresponding magnetic resonance imaging. Beyond image analysis, DL can be used for quality control, workflow organization, and reporting.In this article, we review the most current AI models used in medical imaging research, providing a brief explanation of the various models described in the literature within the past 5 years. Emphasis is placed on the various DL models, as they are the most state-of-art in imaging analysis.

The Artificial Intelligence in Digital Radiology: Part 1: The Challenges, Acceptance and Consensus

“…In line with the first objective of the study, we analysed the main challenges in the design and integration of Artificial Intelligence in digital radiology. The eligibility process led to the choice of 20 papers [ 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 ], among which there are mainly reviews (15 in number) (as it could be expected considering the broad topics covered) but also 5 recent scientific articles/focus articles on very specific aspects [ 15 , 26 , 27 , 29 , 30 ].…”

“…Another study by Maowad et al [ 12 ] specifically focused on the challenges of machine learning and deep learning, subclasses of Artificial Intelligence that showed breakthrough performance in image analysis. The authors discussed the current applications of machine learning and deep learning in the field of diagnostic radiology.…”

“…The first point of view reviewed the challenges and grouped them into six main topics. The first topic is represented by the challenges in the design and employment of the algorithms [ 11 , 12 , 27 ], in their performance improvement [ 11 ], in their potential, not also limited to the imaging field [ 16 ], and their specific application based on the solution [ 27 ]. A second topic was the revolution of the workflow [ 16 ] of the radiologist; many activities could be possible in relation to imaging and many other activities in relation to the administration of other phases of the patient management process (i.e., patient scheduling).…”

Neural Network in the Analysis of the MR Signal as an Image Segmentation Tool for the Determination of T1 and T2 Relaxation Times with Application to Cancer Cell Culture

“…In the literature, work can be found that uses the concept of artificially generated negative data to form decision using a multilayer perceptron [ 35 ]. A few models of neural networks are known, e.g., Convolutional Neural Networks (CNNs) [ 36 , 37 ], Recurrent Neural Networks (RNNs) [ 38 ], Generative Adversarial Networks (GANs) [ 39 ], and Quantitative Susceptibility Mapping (QSM) [ 40 , 41 ]. Particularly, CNNs were most widely used for MRI and other image processing applications.…”

Evolution of the digital operating room: the place of video technology in surgery

“…The recording of the surgical field and the operative environment is increasingly common, with the increase in minimally invasive and robotic surgery necessitating an endoscope. AI through data science has shown to be able to develop the ability to distinguish and diagnose pathology from medical imaging [ 78 ], to recognise anatomical structures from operative footage [ 79 ] and to replicate and automate the performance of surgical tasks [ 80 ]. Regarding the surgical workflow, benefits would be seen with hospital schedules with the ability to predict surgical phases in real time, giving surgical staff a greater awareness of remaining procedural time [ 81 ].…”