Search

Objective There has been a large amount of research in the field of artificial intelligence (AI) as applied to clinical radiology. However, these studies vary in design and quality and systematic reviews of the entire field are lacking.This systematic review aimed to identify all papers that used deep learning in radiology to survey the literature and to evaluate their methods. We aimed to identify the key questions being addressed in the literature and to identify the most effective methods employed. Methods We followed the PRISMA guidelines and performed a systematic review of studies of AI in radiology published from 2015 to 2019. Our published protocol was prospectively registered. Results Our search yielded 11,083 results. Seven hundred sixty-seven full texts were reviewed, and 535 articles were included. Ninety-eight percent were retrospective cohort studies. The median number of patients included was 460. Most studies involved MRI (37%). Neuroradiology was the most common subspecialty. Eighty-eight percent used supervised learning. The majority of studies undertook a segmentation task (39%). Performance comparison was with a state-of-the-art model in 37%. The most used established architecture was UNet (14%). The median performance for the most utilised evaluation metrics was Dice of 0.89 (range .49–.99), AUC of 0.903 (range 1.00–0.61) and Accuracy of 89.4 (range 70.2–100). Of the 77 studies that externally validated their results and allowed for direct comparison, performance on average decreased by 6% at external validation (range increase of 4% to decrease 44%). Conclusion This systematic review has surveyed the major advances in AI as applied to clinical radiology. Key Points • While there are many papers reporting expert-level results by using deep learning in radiology, most apply only a narrow range of techniques to a narrow selection of use cases. • The literature is dominated by retrospective cohort studies with limited external validation with high potential for bias. • The recent advent of AI extensions to systematic reporting guidelines and prospective trial registration along with a focus on external validation and explanations show potential for translation of the hype surrounding AI from code to clinic.

To determine whether cardiac computed tomography (MDCT) can differentiate between functional and valvular aetiologies of chronic mitral regurgitation (MR) compared with echocardiography (TTE).Twenty-seven patients with functional or valvular MR diagnosed by TTE and 19 controls prospectively underwent cardiac MDCT. The morphological appearance of the mitral valve (MV) leaflets, MV geometry, MV leaflet angle, left ventricular (LV) sphericity and global/regional wall motion were analysed. The coronary arteries were evaluated for obstructive atherosclerosis.All control and MR cases were correctly identified by MDCT. Significant differences were detected between valvular and control groups for anterior leaflet length (30 +/- 7 mm vs. 22 +/- 4 mm, P0.02) and thickness (3.0 +/- 1 mm vs. 2.2 +/- 1 mm, P0.01). High-grade coronary stenosis was detected in all patients with functional MR compared with no controls (P0.001). Significant differences in those with/without MV prolapse were detected in MV tent area (-1.0 +/- 0.6 mm vs. 1.3 +/- 0.9 mm, P0.0001) and MV tent height (-0.7 +/- 0.3 mm vs. 0.8 +/- 0.8 mm, P0.0001). Posterior leaflet angle was significantly greater for functional MR (37.9 +/- 19.1 degrees vs. 22.9 +/- 14 degrees , P0.018) and less for valvular MR (0.6 +/- 35.5 degrees vs. 22.9 +/- 14 degrees, P0.017). Sensitivity, specificity, and positive and negative predictive values of MDCT were 100%, 95%, 96% and 100%.Cardiac MDCT allows the differentiation between functional and valvular causes of MR.