Convolutional Neural Networks for Automated Fracture Detection and Localization on Wrist Radiographs

Learning visual models of object categories notoriously requires hundreds or thousands of training examples. We show that it is possible to learn much information about a category from just one, or a handful, of images. The key insight is that, rather than learning from scratch, one can take advantage of knowledge coming from previously learned categories, no matter how different these categories might be. We explore a Bayesian implementation of this idea. Object categories are represented by probabilistic models. Prior knowledge is represented as a probability density function on the parameters of these models. The posterior model for an object category is obtained by updating the prior in the light of one or more observations. We test a simple implementation of our algorithm on a database of 101 diverse object categories. We compare category models learned by an implementation of our Bayesian approach to models learned from by Maximum Likelihood (ML) and Maximum A Posteriori (MAP) methods. We find that on a database of more than 100 categories, the Bayesian approach produces informative models when the number of training examples is too small for other methods to operate successfully.

Background and purpose — We aimed to evaluate the ability of artificial intelligence (a deep learning algorithm) to detect and classify proximal humerus fractures using plain anteroposterior shoulder radiographs. Patients and methods — 1,891 images (1 image per person) of normal shoulders (n = 515) and 4 proximal humerus fracture types (greater tuberosity, 346; surgical neck, 514; 3-part, 269; 4-part, 247) classified by 3 specialists were evaluated. We trained a deep convolutional neural network (CNN) after augmentation of a training dataset. The ability of the CNN, as measured by top-1 accuracy, area under receiver operating characteristics curve (AUC), sensitivity/specificity, and Youden index, in comparison with humans (28 general physicians, 11 general orthopedists, and 19 orthopedists specialized in the shoulder) to detect and classify proximal humerus fractures was evaluated. Results — The CNN showed a high performance of 96% top-1 accuracy, 1.00 AUC, 0.99/0.97 sensitivity/specificity, and 0.97 Youden index for distinguishing normal shoulders from proximal humerus fractures. In addition, the CNN showed promising results with 65–86% top-1 accuracy, 0.90–0.98 AUC, 0.88/0.83–0.97/0.94 sensitivity/specificity, and 0.71–0.90 Youden index for classifying fracture type. When compared with the human groups, the CNN showed superior performance to that of general physicians and orthopedists, similar performance to orthopedists specialized in the shoulder, and the superior performance of the CNN was more marked in complex 3- and 4-part fractures. Interpretation — The use of artificial intelligence can accurately detect and classify proximal humerus fractures on plain shoulder AP radiographs. Further studies are necessary to determine the feasibility of applying artificial intelligence in the clinic and whether its use could improve care and outcomes compared with current orthopedic assessments.

Worldwide interest in artificial intelligence (AI) applications, including imaging, is high and growing rapidly, fueled by availability of large datasets ("big data"), substantial advances in computing power, and new deep-learning algorithms. Apart from developing new AI methods per se, there are many opportunities and challenges for the imaging community, including the development of a common nomenclature, better ways to share image data, and standards for validating AI program use across different imaging platforms and patient populations. AI surveillance programs may help radiologists prioritize work lists by identifying suspicious or positive cases for early review. AI programs can be used to extract "radiomic" information from images not discernible by visual inspection, potentially increasing the diagnostic and prognostic value derived from image datasets. Predictions have been made that suggest AI will put radiologists out of business. This issue has been overstated, and it is much more likely that radiologists will beneficially incorporate AI methods into their practices. Current limitations in availability of technical expertise and even computing power will be resolved over time and can also be addressed by remote access solutions. Success for AI in imaging will be measured by value created: increased diagnostic certainty, faster turnaround, better outcomes for patients, and better quality of work life for radiologists. AI offers a new and promising set of methods for analyzing image data. Radiologists will explore these new pathways and are likely to play a leading role in medical applications of AI.