Abstract
Hepatocellular carcinoma (HCC) is one of the most common malignant tumors and a leading cause of cancer-related death worldwide. We propose a fully automated deep learning model to detect HCC using hepatobiliary phase magnetic resonance images from 549 patients who underwent surgical resection. Our model used a fine-tuned convolutional neural network and achieved 87% sensitivity and 93% specificity for the detection of HCCs with an external validation data set (54 patients). We also confirmed whether the lesion detected by our deep learning model is a true lesion using a class activation map.
Highlights
Deep learning has shown remarkable results in the field of computer vision[9]
External validation using the training generation model for 4,537 images obtained by various MR scanners from multiple vendors showed an 87% sensitivity for Hepatocellular carcinoma (HCC), 93% specificity, and www.nature.com/scientificreports an area under curve (AUC) of 0.90
Our model seems to be more sensitive than less experienced radiologists in detecting very small HCCs
Summary
Deep learning has shown remarkable results in the field of computer vision[9]. Deep learning-based methods have demonstrated that they are well suited for recognition and classification of medical images[10] and they can be used as an effective screening tool in medical image analysis[11]. Deep learning systems can be an auxiliary diagnostic system for the diagnosis of HCC, as well. There are no deep learning-based HCC detection systems using liver MRI in the English literature. The purpose of this study was to develop a fully automated deep learning model to detect HCC using hepatobiliary phase MR images in patients who underwent surgical resection for HCC and evaluate its performance in detecting HCC on liver MRI compared to human readers
Sign-in/Register to view highlights & summary 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.
