Deep learning method with a convolutional neural network for image classification of normal and metastatic axillary lymph nodes on breast ultrasonography.

PURPOSE: To investigate the ability of deep learning (DL) using convolutional neural networks (CNNs) for distinguishing between normal and metastatic axillary lymph nodes on ultrasound images by comparing the diagnostic performance of radiologists. MATERIALS AND METHODS: We retrospectively gathered 300 images of normal and 328 images of axillary lymph nodes with breast cancer metastases for training. A DL model using the CNN architecture Xception was developed to analyze test data of 50 normal and 50 metastatic lymph nodes. A board-certified radiologist with 12 years' experience. (Reader 1) and two residents with 3- and 1-year experience (Readers 2, 3), respectively, scored these test data with and without the assistance of the DL system for the possibility of metastasis. The sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were calculated. RESULTS: Our DL model had a sensitivity of 94%, a specificity of 88%, and an AUC of 0.966. The AUC of the DL model was not significantly different from that of Reader 1 (0.969; p = 0.881) and higher than that of Reader 2 (0.913; p = 0.101) and Reader 3 (0.810; p < 0.001). With the DL support, the AUCs of Readers 2 and 3 increased to 0.960 and 0.937, respectively, which were comparable to those of Reader 1 (p = 0.138 and 0.700, respectively). CONCLUSION: Our DL model demonstrated great diagnostic performance for differentiating benign from malignant axillary lymph nodes on breast ultrasound and for potentially providing effective diagnostic support to residents.

Deep Learning Using Multiple Degrees of Maximum-Intensity Projection for PET/CT Image Classification in Breast Cancer.

Deep learning (DL) has become a remarkably powerful tool for image processing recently. However, the usefulness of DL in positron emission tomography (PET)/computed tomography (CT) for breast cancer (BC) has been insufficiently studied. This study investigated whether a DL model using images with multiple degrees of PET maximum-intensity projection (MIP) images contributes to increase diagnostic accuracy for PET/CT image classification in BC. We retrospectively gathered 400 images of 200 BC and 200 non-BC patients for training data. For each image, we obtained PET MIP images with four different degrees (0°, 30°, 60°, 90°) and made two DL models using Xception. One DL model diagnosed BC with only 0-degree MIP and the other used four different degrees. After training phases, our DL models analyzed test data including 50 BC and 50 non-BC patients. Five radiologists interpreted these test data. Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were calculated. Our 4-degree model, 0-degree model, and radiologists had a sensitivity of 96%, 82%, and 80-98% and a specificity of 80%, 88%, and 76-92%, respectively. Our 4-degree model had equal or better diagnostic performance compared with that of the radiologists (AUC = 0.936 and 0.872-0.967, p = 0.036-0.405). A DL model similar to our 4-degree model may lead to help radiologists in their diagnostic work in the future.