Xin Li1, Genggeng Qin2, Qiang He1, Lei Sun1, Hui Zeng2, Zilong He2, Weiguo Chen2, Xin Zhen3, Linghong Zhou4. 1. School of Biomedical Engineering, Southern Medical University, Guangzhou, 510515, Guangdong, China. 2. Department of Radiology, Nanfang Hospital, Southern Medical University, Guangzhou, 510515, Guangdong, China. 3. School of Biomedical Engineering, Southern Medical University, Guangzhou, 510515, Guangdong, China. xinzhen@smu.edu.cn. 4. School of Biomedical Engineering, Southern Medical University, Guangzhou, 510515, Guangdong, China. smart@smu.edu.cn.
Abstract
OBJECTIVE: To evaluate the impact of utilizing digital breast tomosynthesis (DBT) or/and full-field digital mammography (FFDM), and different transfer learning strategies on deep convolutional neural network (DCNN)-based mass classification for breast cancer. METHODS: We retrospectively collected 441 patients with both DBT and FFDM on which regions of interest (ROIs) covering the malignant, benign and normal tissues were extracted for DCNN training and validation. Experiments were conducted for tasks in distinguishing malignant/benign/normal: (1) classification capabilities of DBT vs FFDM and the role of transfer learning were validated on 2D-DCNN; (2) different strategies of combining DBT and FFDM and the associated impacts on classification were explored; (3) 2D-DCNN and 3D-DCNN trained from scratch with volumetric DBT were compared. RESULTS: 2D-DCNN with transfer learning outperformed that without for DBT in distinguishing malignant (ΔAUC = 0.059 ± 0.009, p < 0.001), benign (ΔAUC = 0.095 ± 0.010, p < 0.001) and normal tissue (ΔAUC = 0.042 ± 0.004, p < 0.001) (paired samples t test). 2D-DCNN trained on DBT (with transfer learning) achieved higher accuracy than those on FFDM (malignant: ΔAUC = 0.014 ± 0.014, p = 0.037; benign: ΔAUC = 0.031 ± 0.006, p < 0.001; normal: ΔAUC = 0.017 ± 0.004, p < 0.001) (independent samples t test). The 2D-DCNN employing both DBT and FFDM for training achieved better performances in benign (FFDM: ΔAUC = 0.010 ± 0.008, p < 0.001; DBT: ΔAUC = 0.009 ± 0.005, p < 0.001) and normal (FFDM: ΔAUC = 0.005 ± 0.003, p < 0.001; DBT: ΔAUC = 0.002 ± 0.002, p < 0.001) (related samples Friedman test). The 3D-DCNN and 2D-DCNN trained from scratch with DBT only produced moderate classification. CONCLUSIONS: Transfer learning facilitates mass classification for both DBT and FFDM, and DBT outperforms FFDM when equipped with transfer learning. Integrating DBT and FFDM in DCNN training enhances mass classification accuracy for breast cancer. KEY POINTS: • Transfer learning facilitates mass classification for both DBT and FFDM, and the DBT-based DCNN outperforms the FFDM-based DCNN when equipped with transfer learning. • Integrating DBT and FFDM in DCNN training enhances breast mass classification accuracy. • 3D-DCNN/2D-DCNN trained from scratch with volumetric DBT but without transfer learning only produce moderate mass classification result.
OBJECTIVE: To evaluate the impact of utilizing digital breast tomosynthesis (DBT) or/and full-field digital mammography (FFDM), and different transfer learning strategies on deep convolutional neural network (DCNN)-based mass classification for breast cancer. METHODS: We retrospectively collected 441 patients with both DBT and FFDM on which regions of interest (ROIs) covering the malignant, benign and normal tissues were extracted for DCNN training and validation. Experiments were conducted for tasks in distinguishing malignant/benign/normal: (1) classification capabilities of DBT vs FFDM and the role of transfer learning were validated on 2D-DCNN; (2) different strategies of combining DBT and FFDM and the associated impacts on classification were explored; (3) 2D-DCNN and 3D-DCNN trained from scratch with volumetric DBT were compared. RESULTS: 2D-DCNN with transfer learning outperformed that without for DBT in distinguishing malignant (ΔAUC = 0.059 ± 0.009, p < 0.001), benign (ΔAUC = 0.095 ± 0.010, p < 0.001) and normal tissue (ΔAUC = 0.042 ± 0.004, p < 0.001) (paired samples t test). 2D-DCNN trained on DBT (with transfer learning) achieved higher accuracy than those on FFDM (malignant: ΔAUC = 0.014 ± 0.014, p = 0.037; benign: ΔAUC = 0.031 ± 0.006, p < 0.001; normal: ΔAUC = 0.017 ± 0.004, p < 0.001) (independent samples t test). The 2D-DCNN employing both DBT and FFDM for training achieved better performances in benign (FFDM: ΔAUC = 0.010 ± 0.008, p < 0.001; DBT: ΔAUC = 0.009 ± 0.005, p < 0.001) and normal (FFDM: ΔAUC = 0.005 ± 0.003, p < 0.001; DBT: ΔAUC = 0.002 ± 0.002, p < 0.001) (related samples Friedman test). The 3D-DCNN and 2D-DCNN trained from scratch with DBT only produced moderate classification. CONCLUSIONS: Transfer learning facilitates mass classification for both DBT and FFDM, and DBT outperforms FFDM when equipped with transfer learning. Integrating DBT and FFDM in DCNN training enhances mass classification accuracy for breast cancer. KEY POINTS: • Transfer learning facilitates mass classification for both DBT and FFDM, and the DBT-based DCNN outperforms the FFDM-based DCNN when equipped with transfer learning. • Integrating DBT and FFDM in DCNN training enhances breast mass classification accuracy. • 3D-DCNN/2D-DCNN trained from scratch with volumetric DBT but without transfer learning only produce moderate mass classification result.
Entities:
Keywords:
Breast; Classification; Deep learning; Mammography; Neural network (computer)
Authors: Bo Kyoung Seo; Etta D Pisano; Cherie M Kuzmiak; Marcia Koomen; Dag Pavic; Robert McLelland; Yeonhee Lee; Elodia B Cole; Dianne Mattingly; Juneyoung Lee Journal: Acad Radiol Date: 2006-10 Impact factor: 3.173
Authors: S Mall; J Noakes; M Kossoff; W Lee; M McKessar; A Goy; J Duncombe; M Roberts; B Giuffre; A Miller; N Bhola; C Kapoor; C Shearman; G DaCosta; S Choi; J Sterba; M Kay; K Bruderlin; N Winarta; K Donohue; B Macdonell-Scott; F Klijnsma; K Suzuki; P Brennan; C Mello-Thoms Journal: Eur Radiol Date: 2018-05-30 Impact factor: 5.315
Authors: Melissa A Durand; Brian M Haas; Xiaopan Yao; Jaime L Geisel; Madhavi Raghu; Regina J Hooley; Laura J Horvath; Liane E Philpotts Journal: Radiology Date: 2014-09-01 Impact factor: 11.105
Authors: Constance D Lehman; Robert F Arao; Brian L Sprague; Janie M Lee; Diana S M Buist; Karla Kerlikowske; Louise M Henderson; Tracy Onega; Anna N A Tosteson; Garth H Rauscher; Diana L Miglioretti Journal: Radiology Date: 2016-12-05 Impact factor: 11.105
Authors: John R Zech; Marcus A Badgeley; Manway Liu; Anthony B Costa; Joseph J Titano; Eric Karl Oermann Journal: PLoS Med Date: 2018-11-06 Impact factor: 11.069