Differential Diagnosis of DCIS and Fibroadenoma Based on Ultrasound Images: a Difference-Based Self-Supervised Approach

Jin Yin1,2, Jia-Jun Qiu1, Jing-Yan Liu3, Yi-Yue Li1, Qi-Cheng Lao4, Xiao-Rong Zhong5, Mengling Feng6, Hao Du6, Shao-Liang Peng7, Yu-Lan Peng3
1West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, China
2School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, China
3Department of Ultrasonography, West China Hospital, Sichuan University, Chengdu, China
4School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing, China
5Breast Disease Center, West China Hospital, Sichuan University Cancer Center, Chengdu, China
6Saw Swee Hock School of Public Health, National University Health System, National University of Singapore, Singapore, Singapore
7College of Computer Science and Electronic Engineering, Hunan University, Changsha, China

Tóm tắt

Differentiation of ductal carcinoma in situ (DCIS, a precancerous lesion of the breast) from fibroadenoma (FA) using ultrasonography is significant for the early prevention of malignant breast tumors. Radiomics-based artificial intelligence (AI) can provide additional diagnostic information but usually requires extensive labeling efforts by clinicians with specialized knowledge. This study aims to investigate the feasibility of differentially diagnosing DCIS and FA using ultrasound radiomics-based AI techniques and further explore a novel approach that can reduce labeling efforts without sacrificing diagnostic performance. We included 461 DCIS and 651 FA patients, of whom 139 DCIS and 181 FA patients constituted a prospective test cohort. First, various feature engineering-based machine learning (FEML) and deep learning (DL) approaches were developed. Then, we designed a difference-based self-supervised (DSS) learning approach that only required FA samples to participate in training. The DSS approach consists of three steps: (1) pretraining a Bootstrap Your Own Latent (BYOL) model using FA images, (2) reconstructing images using the encoder and decoder of the pretrained model, and (3) distinguishing DCIS from FA based on the differences between the original and reconstructed images. The experimental results showed that the trained FEML and DL models achieved the highest AUC of 0.7935 (95% confidence interval, 0.7900–0.7969) on the prospective test cohort, indicating that the developed models are effective for assisting in differentiating DCIS from FA based on ultrasound images. Furthermore, the DSS model achieved an AUC of 0.8172 (95% confidence interval, 0.8124–0.8219), indicating that our model outperforms the conventional radiomics-based AI models and is more competitive.

Tài liệu tham khảo

Afshar P, Mohammadi A, Plataniotis KN, Oikonomou A, Benali H (2019) From handcrafted to deep-learning-based cancer radiomics: challenges and opportunities. IEEE Signal Process Mag 36(4):132–160. https://doi.org/10.1109/msp.2019.2900993 Boyd S, Parikh N, Chu E, Peleato B, Eckstein J (2010) Distributed optimization and statistical learning via the alternating direction method of multipliers. Found Trend Mach Learn 3(1):1–122. https://doi.org/10.1561/2200000016 Ding C, Peng H (2005) Minimum redundancy feature selection from microarray gene expression data. J Bioinform Comput Biol 3(2):185–205. https://doi.org/10.1142/s0219720005001004 Erlichman DB, Weiss A, Koenigsberg M, Stein MW (2020) Contrast enhanced ultrasound: a review of radiology applications. Clin Imaging 60(2):209–215. https://doi.org/10.1016/j.clinimag.2019.12.013 Fleury EDFC, Marcomini K (2020) Impact of radiomics on the breast ultrasound radiologist’s clinical practice: from lumpologist to data wrangler. Europ J Radiol. 131:109197. https://doi.org/10.1016/j.ejrad.2020.109197 Friedman J, Hastie T, Tibshirani R (2010) Regularization paths for generalized linear models via coordinate descent. J Statis Software. 33(1):1–22 Fujioka T, Kubota K, Mori M, Kikuchi Y, Katsuta L, Kasahara M, Oda G, Ishiba T, Nakagawa T, Tateishi U (2019) Distinction between benign and malignant breast masses at breast ultrasound using deep learning method with convolutional neural network. Jpn J Radiol 37(6):466–472. https://doi.org/10.1007/s11604-019-00831-5 Gillies RJ, Kinahan PE, Hricak H (2016) Radiomics: images are more than pictures, they are data. Radiology 278(2):563–577. https://doi.org/10.1148/radiol.2015151169 Grill J-B, Strub F, Altché F, Tallec C, Richemond P, Buchatskaya E, Doersch C, Avila Pires B, Guo Z, Gheshlaghi Azar M (2020) Bootstrap your own latent-a new approach to self-supervised learning. Adv Neural Inf Process Syst 33:21271–21284 Gruber R, Jaromi S, Rudas M, Pfarl G, Riedl CC, Flory D, Graf O, Sickles EA, Helbich TH (2013) Histologic work-up of non-palpable breast lesions classified as probably benign at initial mammography and/or ultrasound (BI-RADS category 3). Eur J Radiol 82(3):398–403. https://doi.org/10.1016/j.ejrad.2012.02.004 Hanchuan P, Fuhui L, Ding C (2005) Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans Pattern Anal Mach Intell 27(8):1226–1238. https://doi.org/10.1109/TPAMI.2005.159 He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. IEEE Conf Compt Vision Patt Recogn. https://doi.org/10.1109/CVPR.2016.90 Iandola FN, Han S, Moskewicz MW, Ashraf K, Dally WJ, Keutzer K (2016) SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 05 MB model size. arXiv preprint arXiv. 1602.07360. Krizhevsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep convolutional neural networks. Adv Neural Inf Process Syst 60(6):84–90. https://doi.org/10.1145/3065386 Kubota K (2021) Breast ultrasound in the age of advanced technology and artificial intelligence. J Med Ultrason 48(2):113–114. https://doi.org/10.1007/s10396-021-01091-5 Lambin P, Leijenaar RTH, Deist TM, Peerlings J, De Jong EEC, Van Timmeren J, Sanduleanu S, Larue RTHM, Even AJG, Jochems A, Van Wijk Y, Woodruff H, Van Soest J, Lustberg T, Roelofs E, Van Elmpt W, Dekker A, Mottaghy FM, Wildberger JE, Walsh S (2017) Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev Clin Oncol 14(12):749–762. https://doi.org/10.1038/nrclinonc.2017.141 Lee SE, Han K, Kwak JY, Lee E, Kim E-K (2018) Radiomics of US texture features in differential diagnosis between triple-negative breast cancer and fibroadenoma. Sci Rep 8(1):13546. https://doi.org/10.1038/s41598-018-31906-4 Li W, Zhou Q, Xia S, Wu Y, Fei X, Wang Y, Tao L, Fan J, Zhou W (2020) Application of contrast-enhanced ultrasound in the diagnosis of ductal carcinoma in situ: analysis of 127 cases. J Ultrasound Med 39(1):39–50. https://doi.org/10.1002/jum.15069 Litjens G, Kooi T, Bejnordi BE, Setio Aa A, Ciompi F, Ghafoorian M, Van Der Laak Ja WM, Van Ginneken B, Sanchez CI (2017) A survey on deep learning in medical image analysis. Med Image Analy. 42:60–88. https://doi.org/10.1016/j.media.2017.07.005 Liu X, Zhang F, Hou Z, Mian L, Wang Z, Zhang J, Tang J (2021) Self-supervised learning: generative or contrastive. IEEE Trans Knowl Data Eng. https://doi.org/10.1109/TKDE.2021.3090866 Magnuska ZA, Theek B, Darguzyte M, Palmowski M, Stickeler E, Schulz V, Kiessling F (2022) Influence of the computer-aided decision support system design on ultrasound-based breast cancer classification. Cancers. https://doi.org/10.3390/cancers14020277 Moon WK, Huang Y-S, Lo C-M, Huang C-S, Bae MS, Kim WH, Chen J-H, Chang R-F (2015) Computer-aided diagnosis for distinguishing between triple-negative breast cancer and fibroadenomas based on ultrasound texture features. Med Phys 42(6):3024–3035. https://doi.org/10.1118/1.4921123 O’connell AM, Bartolotta TV, Orlando A, Jung S-H, Baek J, Parker KJ (2022) Diagnostic performance of an artificial intelligence system in breast ultrasound. J Ultrasound Med 41(1):97–105. https://doi.org/10.1002/jum.15684 Qiu J-J, Yin J, Ji L, Lu C-Y, Li K, Zhang Y-G, Lin Y-XJIP (2022) Management Differential diagnosis of hepatocellular carcinoma and hepatic hemangioma based on maximum wavelet-coefficient statistics: novel radiomics features from plain CT. Inform Process Manag. 59(5):103046 Qiu J-J, Yin J, Qian W, Liu J-H, Huang Z-X, Yu H-P, Ji L, Zeng X-X (2021) A novel multiresolution-statistical texture analysis architecture radiomics-aided diagnosis of pdac based on plain CT images. IEEE Trans Med Imaging 40(1):12–25. https://doi.org/10.1109/TMI.2020.3021254 Redmon J, Farhadi A (2018) Yolov3: An incremental improvement. arXiv preprint arXiv. 1804.02767. Roffo G, Melzi S, Castellani U, Vinciarelli A (2017) Infinite latent feature selection: a probabilistic latent graph-based ranking approach. International Conference on Computer Vision. Venice, Italy. Russakovsky O, Deng J, Su H, Krause J, Satheesh S, Ma S, Huang Z, Karpathy A, Khosla A, Bernstein M, Berg AC, Fei-Fei L (2015) ImageNet large scale visual recognition challenge. Int J Comput Vision 115(3):211–252. https://doi.org/10.1007/s11263-015-0816-y Shen D, Wu G, Suk H-I (2017) Deep Learning in medical image analysis. Annu Rev Biomed Eng 19:221–248. https://doi.org/10.1146/annurev-bioeng-071516-044442 Shen Y, Shamout FE, Oliver JR, Witowski J, Kannan K, Park J, Wu N, Huddleston C, Wolfson S, Millet A, Ehrenpreis R, Awal D, Tyma C, Samreen N, Gao Y, Chhor C, Gandhi S, Lee C, Kumari-Subaiya S, Leonard C, Mohammed R, Moczulski C, Altabet J, Babb J, Lewin A, Reig B, Moy L, Heacock L, Geras KJ (2021) Artificial intelligence system reduces false-positive findings in the interpretation of breast ultrasound exams. Nat Commun 12(1):5645. https://doi.org/10.1038/s41467-021-26023-2 Shia W-C, Chen D-R (2021) Classification of malignant tumors in breast ultrasound using a pretrained deep residual network model and support vector machine. Compt Med Imag Graph. 87:101829. https://doi.org/10.1016/j.compmedimag.2020.101829 Sim Y, Lee SE, Kim E-K, Kim S (2020) A radiomics approach for the classification of fibroepithelial lesions on breast ultrasonography. Ultrasound Med Biol 46(5):1133–1141. https://doi.org/10.1016/j.ultrasmedbio.2020.01.015 Sinn H-P, Kreipe H (2013) A brief overview of the who classification of breast tumors, 4th edition, focusing on issues and updates from the 3rd edition. Breast Care. 8(2):149–154. https://doi.org/10.1159/000350774 Strzelecki M, Szczypinski P, Materka A, Klepaczko A (2013) A software tool for automatic classification and segmentation of 2D/3D medical images. Nucl Instrum Methods Phys Res, Sect A 702:137–140. https://doi.org/10.1016/j.nima.2012.09.006 Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z (2016) Rethinking the inception architecture for computer vision. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Tan M, Le Q (2019) Efficientnet: Rethinking model scaling for convolutional neural networks. International Conference on Machine Learning. Tibshirani R (1996) Regression shrinkage and selection via the lasso. J Roy Stat Soc: Ser B (Methodol) 58(1):267–288. https://doi.org/10.1111/j.2517-6161.1996.tb02080.x Van Griethuysen JJM, Fedorov A, Parmar C, Hosny A, Aucoin N, Narayan V, Beets-Tan RGH, Fillion-Robin JC, Pieper S, Aerts HJWL (2017) Computational radiomics system to decode the radiographic phenotype. Can Res 77(21):e104–e107. https://doi.org/10.1158/0008-5472.CAN-17-0339 Wang X-Y, Cui L-G, Feng J, Chen W (2021) Artificial intelligence for breast ultrasound: an adjunct tool to reduce excessive lesion biopsy. Europ J Radiol. 138:109624. https://doi.org/10.1016/j.ejrad.2021.109624 Xu M, Li F, Yu S, Zeng S, Weng G, Teng P, Yang H, Li X, Liu G (2022) Value of histogram of gray-scale ultrasound image in differential diagnosis of small triple negative breast invasive ductal carcinoma and fibroadenoma. Cancer Manage Res 14:1515–1524. https://doi.org/10.2147/CMAR.S359986 Zhang X, Lin X, Zhang Z, Dong L, Sun X, Sun D, Yuan K (2020) Artificial intelligence medical ultrasound equipment: application of breast lesions detection. Ultrason Imaging 42(4–5):191–202. https://doi.org/10.1177/0161734620928453