Yang Zhang1, Siwa Chan2, Vivian Youngjean Park3, Kai-Ting Chang1, Siddharth Mehta1, Min Jung Kim3, Freddie J Combs1, Peter Chang1, Daniel Chow1, Ritesh Parajuli4, Rita S Mehta4, Chin-Yao Lin2, Sou-Hsin Chien2, Jeon-Hor Chen5, Min-Ying Su6. 1. Department of Radiological Sciences, University of California, 164 Irvine Hall, Irvine, CA 92697-5020. 2. Department of Medical Imaging, Taichung Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Taichung, Taiwan; School of Medicine, Tzu Chi University, Hualien, Taiwan. 3. Department of Radiology and Research Institute of Radiological Science, Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea. 4. Department of Medicine, University of California, Irvine, California. 5. Department of Radiological Sciences, University of California, 164 Irvine Hall, Irvine, CA 92697-5020; Department of Radiology, E-Da Hospital and I-Shou University, Kaohsiung, Taiwan. 6. Department of Radiological Sciences, University of California, 164 Irvine Hall, Irvine, CA 92697-5020. Electronic address: msu@uci.edu.
Abstract
RATIONALE AND OBJECTIVES: Computer-aided methods have been widely applied to diagnose lesions on breast magnetic resonance imaging (MRI). The first step was to identify abnormal areas. A deep learning Mask Regional Convolutional Neural Network (R-CNN) was implemented to search the entire set of images and detect suspicious lesions. MATERIALS AND METHODS: Two DCE-MRI datasets were used, 241 patients acquired using non-fat-sat sequence for training, and 98 patients acquired using fat-sat sequence for testing. All patients have confirmed unilateral mass cancers. The tumor was segmented using fuzzy c-means clustering algorithm to serve as the ground truth. Mask R-CNN was implemented with ResNet-101 as the backbone. The neural network output the bounding boxes and the segmented tumor for evaluation using the Dice Similarity Coefficient (DSC). The detection performance, and the trade-off between sensitivity and specificity, was analyzed using free response receiver operating characteristic. RESULTS: When the precontrast and subtraction image of both breasts were used as input, the false positive from the heart and normal parenchymal enhancements could be minimized. The training set had 1469 positive slices (containing lesion) and 9135 negative slices. In 10-fold cross-validation, the mean accuracy = 0.86 and DSC = 0.82. The testing dataset had 1568 positive and 7264 negative slices, with accuracy = 0.75 and DSC = 0.79. When the obtained per-slice results were combined, 240 of 241 (99.5%) lesions in the training and 98 of 98 (100%) lesions in the testing datasets were identified. CONCLUSION: Deep learning using Mask R-CNN provided a feasible method to search breast MRI, localize, and segment lesions. This may be integrated with other artificial intelligence algorithms to develop a fully automatic breast MRI diagnostic system.
RATIONALE AND OBJECTIVES: Computer-aided methods have been widely applied to diagnose lesions on breast magnetic resonance imaging (MRI). The first step was to identify abnormal areas. A deep learning Mask Regional Convolutional Neural Network (R-CNN) was implemented to search the entire set of images and detect suspicious lesions. MATERIALS AND METHODS: Two DCE-MRI datasets were used, 241 patients acquired using non-fat-sat sequence for training, and 98 patients acquired using fat-sat sequence for testing. All patients have confirmed unilateral mass cancers. The tumor was segmented using fuzzy c-means clustering algorithm to serve as the ground truth. Mask R-CNN was implemented with ResNet-101 as the backbone. The neural network output the bounding boxes and the segmented tumor for evaluation using the Dice Similarity Coefficient (DSC). The detection performance, and the trade-off between sensitivity and specificity, was analyzed using free response receiver operating characteristic. RESULTS: When the precontrast and subtraction image of both breasts were used as input, the false positive from the heart and normal parenchymal enhancements could be minimized. The training set had 1469 positive slices (containing lesion) and 9135 negative slices. In 10-fold cross-validation, the mean accuracy = 0.86 and DSC = 0.82. The testing dataset had 1568 positive and 7264 negative slices, with accuracy = 0.75 and DSC = 0.79. When the obtained per-slice results were combined, 240 of 241 (99.5%) lesions in the training and 98 of 98 (100%) lesions in the testing datasets were identified. CONCLUSION: Deep learning using Mask R-CNN provided a feasible method to search breast MRI, localize, and segment lesions. This may be integrated with other artificial intelligence algorithms to develop a fully automatic breast MRI diagnostic system.
Authors: Elise Marechal; Adrien Jaugey; Georges Tarris; Michel Paindavoine; Jean Seibel; Laurent Martin; Mathilde Funes de la Vega; Thomas Crepin; Didier Ducloux; Gilbert Zanetta; Sophie Felix; Pierre Henri Bonnot; Florian Bardet; Luc Cormier; Jean-Michel Rebibou; Mathieu Legendre Journal: Clin J Am Soc Nephrol Date: 2021-12-03 Impact factor: 8.237
Authors: John A Cole; Rita Nanda; Frederick M Howard; Gong He; Joseph R Peterson; J R Pfeiffer; Tyler Earnest; Alexander T Pearson; Hiroyuki Abe Journal: Breast Cancer Res Treat Date: 2022-09-05 Impact factor: 4.624