A deep supervised transformer U‐shaped full‐resolution residual network for the segmentation of breast ultrasound image

Abstract

PurposeBreast ultrasound (BUS) is an important breast imaging tool. Automatic BUS image segmentation can measure the breast tumor size objectively and reduce doctors’ workload. In this article, we proposed a deep supervised transformer U‐shaped full‐resolution residual network (DSTransUFRRN) to segment BUS images.MethodsIn the proposed method, a full‐resolution residual stream and a deep supervision mechanism were introduced into TransU‐Net. The residual stream can keep full resolution features from different levels and enhance features fusion. Then, the deep supervision can suppress gradient dispersion. Moreover, the transformer module can suppress irrelevant features and improve feature extraction process. Two datasets (dataset A and B) were used for training and evaluation. The dataset A included 980 BUS image samples and the dataset B had 163 BUS image samples.ResultsCross‐validation was conducted. For the dataset A, the proposed DSTransUFRRN achieved significantly higher Dice (91.04 ± 0.86%) than all compared methods (p < 0.05). For the dataset B, the Dice was lower than that for the dataset A due to the small number of samples, but the Dice of DSTransUFRRN (88.15% ± 2.11%) was significantly higher than that of other compared methods (p < 0.05).ConclusionsIn this study, we proposed DSTransUFRRN for BUS image segmentation. The proposed methods achieved significantly higher accuracy than the compared previous methods.

Ultrasound Image Analysis with Vision Transformers—Review

“…This arises from the quadratic time and space complexity associated with the attention mechanism within the transformer architecture. For instance, U-Net++ models [ 94 ], which are based on CNNs, require approximately 9.163 million parameters to achieve a Dice score of 76.40 on the BUSI dataset [ 38 ]. In contrast, TransUnet [ 46 ], which secures a higher Dice score of 81.18 on the BUSI dataset, necessitates only about 44.00 million parameters [ 38 ].…”

“…For instance, U-Net++ models [ 94 ], which are based on CNNs, require approximately 9.163 million parameters to achieve a Dice score of 76.40 on the BUSI dataset [ 38 ]. In contrast, TransUnet [ 46 ], which secures a higher Dice score of 81.18 on the BUSI dataset, necessitates only about 44.00 million parameters [ 38 ]. Nevertheless, researchers must grapple with the intense demand for GPU resources to meet these demands.…”

“…He et al [ 38 ] introduced a hybrid CNN–transformer network (HCTNet) consisting of transformer encoder blocks (TEBlocks) in the encoder and a spatial-wise cross attention (SCA) module in the decoder to enhance breast lesion segmentation in BUS ultrasound images. Their application of the HCT network highlighted the importance of local features due to a unique computer kernel, though this focus led to difficulties in evaluating tumor-like shadows and speckle noise.…”

Ultrasound Image Analysis with Vision Transformers – Review

“…He et al [38] introduced a hybrid CNN-transformer network (HCTNet) consisting of transformer encoder blocks (TEBlocks) in the encoder and a spatial-wise cross attention (SCA) module in the decoder to enhance breast lesion segmentation in BUS ultrasound images. Their application of the HCT network highlighted the importance of local features due to a unique computer kernel, though this focus led to difficulties in evaluating tumorlike shadows and speckle noise.…”