Abstract
The development of deep learning algorithms for complex tasks in digital medicine has relied on the availability of large labeled training datasets, usually containing hundreds of thousands of examples. The purpose of this study was to develop a 3D deep learning model, AppendiXNet, to detect appendicitis, one of the most common life-threatening abdominal emergencies, using a small training dataset of less than 500 training CT exams. We explored whether pretraining the model on a large collection of natural videos would improve the performance of the model over training the model from scratch. AppendiXNet was pretrained on a large collection of YouTube videos called Kinetics, consisting of approximately 500,000 video clips and annotated for one of 600 human action classes, and then fine-tuned on a small dataset of 438 CT scans annotated for appendicitis. We found that pretraining the 3D model on natural videos significantly improved the performance of the model from an AUC of 0.724 (95% CI 0.625, 0.823) to 0.810 (95% CI 0.725, 0.895). The application of deep learning to detect abnormalities on CT examinations using video pretraining could generalize effectively to other challenging cross-sectional medical imaging tasks when training data is limited.
Highlights
Recent advancements in deep learning have enabled algorithms to automate a wide variety of medical tasks[14,15,16,17,18,19,20]
We developed a deep learning model capable of detecting appendicitis on abdominal CT using a small dataset of 438 CT examinations
The training set consisted of 438 exams (255 appendicitis and 183 non-appendicitis from 435 patients), the development set consisted of 106 exams (53 appendicitis and 53 non-appendicitis from 105 patients), and the test set consisted of 102 exams (51 appendicitis and 51 non-appendicitis from 102 patients)
Summary
Recent advancements in deep learning have enabled algorithms to automate a wide variety of medical tasks[14,15,16,17,18,19,20]. A key aspect for the success of deep learning models on these tasks is the availability of large labeled datasets of medical images[21], usually containing hundreds of thousands of examples. Pretraining[22] can mitigate this problem, but pretraining employs large datasets of natural images such as ImageNet. Pretraining[22] can mitigate this problem, but pretraining employs large datasets of natural images such as ImageNet Such training has been effective for 2D medical imaging tasks (where datasets are comparatively smaller), but does not apply to 3D data produced by cross sectional imaging devices. Successful application of deep learning to detect abnormalities on CT examinations using video pretraining could generalize effectively to other cross-sectional medical imaging deep learning applications where training data is limited and the task is challenging
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
