An attention‐supervised full‐resolution residual network for the segmentation of breast ultrasound images. Issue 11 (6th October 2020)

• Record Type: Journal Article
• Authors: Qu, Xiaolei; Shi, Yao; Hou, Yaxin; Jiang, Jue
• Is Part Of: Medical physics. Volume 47:Issue 11(2020)
• Page Start: 5702
• Page End: 5714
• Publication Date: 2020-10-06
• Subjects: breast cancer -- breast ultrasound image -- deep learning -- segmentation; Medical physics -- Periodicals; Medical physics; Geneeskunde; Natuurkunde; Toepassingen; Biophysics; Periodicals; Periodicals; Electronic journals; 610.153
• DOI: 10.1002/mp.14470
• Languages: English
• ISSNs: 0094-2405

Abstract

Abstract : Purpose: Breast cancer is the most common cancer among women worldwide. Medical ultrasound imaging is one of the widely applied breast imaging methods for breast tumors. Automatic breast ultrasound (BUS) image segmentation can measure the size of tumors objectively. However, various ultrasound artifacts hinder segmentation. We proposed an attention‐supervised full‐resolution residual network (ASFRRN) to segment tumors from BUS images. Methods: In the proposed method, Global Attention Upsample (GAU) and deep supervision were introduced into a full‐resolution residual network (FRRN), where GAU learns to merge features at different levels with attention for deep supervision. Two datasets were employed for evaluation. One (Dataset A) consisted of 163 BUS images with tumors (53 malignant and 110 benign) from UDIAT Centre Diagnostic, and the other (Dataset B) included 980 BUS images with tumors (595 malignant and 385 benign) from the Sun Yat‐sen University Cancer Center. The tumors from both datasets were manually segmented by medical doctors. For evaluation, the Dice coefficient (Dice), Jaccard similarity coefficient (JSC), and F1 score were calculated. Results: For Dataset A, the proposed method achieved higher Dice (84.3 ± 10.0%), JSC (75.2 ± 10.7%), and F1 score (84.3 ± 10.0%) than the previous best method: FRRN. For Dataset B, the proposed method also achieved higher Dice (90.7 ± 13.0%), JSC (83.7 ± 14.8%), and F1 score (90.7 ± 13.0%) than the previous best methods: DeepLabv3 and dual attention network (DANet). For Dataset A + B, the proposed method achieved higher Dice (90.5 ± 13.1%), JSC (83.3 ± 14.8%), and F1 score (90.5 ± 13.1%) than the previous best method: DeepLabv3. Additionally, the parameter number of ASFRRN was only 10.6 M, which is less than those of DANet (71.4 M) and DeepLabv3 (41.3 M). Conclusions: We proposed ASFRRN, which combined with FRRN, attention mechanism, and deep supervision to segment tumors from BUS images. It achieved high segmentation accuracy with a reduced parameter number.