School of Computer Science and Artificial Intelligence, Changzhou University, Changzhou, Jiangsu 213164, China; The Affiliated Changzhou NO.2 People's Hospital of Nanjing Medical University, Changzhou, Jiangsu 213003, China; Key Laboratory of Computer Network and Information Integration, Southeast University, Nanjing, Jiangsu 211096, China.
School of Computer Science and Artificial Intelligence, Changzhou University, Changzhou, Jiangsu 213164, China.
Comput Methods Programs Biomed. 2023 Aug;238:107614. doi: 10.1016/j.cmpb.2023.107614. Epub 2023 May 19.
Accurate and efficient segmentation of thyroid nodules on ultrasound images is critical for computer-aided nodule diagnosis and treatment. For ultrasound images, Convolutional neural networks (CNNs) and Transformers, which are widely used in natural images, cannot obtain satisfactory segmentation results, because they either cannot obtain precise boundaries or segment small objects.
To address these issues, we propose a novel Boundary-preserving assembly Transformer UNet (BPAT-UNet) for ultrasound thyroid nodule segmentation. In the proposed network, a Boundary point supervision module (BPSM), which adopts two novel self-attention pooling approaches, is designed to enhance boundary features and generate ideal boundary points through a novel method. Meanwhile, an Adaptive multi-scale feature fusion module (AMFFM) is constructed to fuse features and channel information at different scales. Finally, to fully integrate the characteristics of high-frequency local and low-frequency global, the Assembled transformer module (ATM) is placed at the bottleneck of the network. The correlation between deformable features and features-among computation is characterized by introducing them into the above two modules of AMFFM and ATM. As the design goal and eventually demonstrated, BPSM and ATM promote the proposed BPAT-UNet to further constrain boundaries, whereas AMFFM assists to detect small objects.
Compared to other classical segmentation networks, the proposed BPAT-UNet displays superior segmentation performance in visualization results and evaluation metrics. Significant improvement of segmentation accuracy was shown on the public thyroid dataset of TN3k with Dice similarity coefficient (DSC) of 81.64% and 95th percentage of the asymmetric Hausdorff distance (HD95) of 14.06, whereas those on our private dataset were with DSC of 85.63% and HD95 of 14.53, respectively.
This paper presents a method for thyroid ultrasound image segmentation, which achieves high accuracy and meets the clinical requirements. Code is available at https://github.com/ccjcv/BPAT-UNet.
在超声图像上准确、高效地分割甲状腺结节对计算机辅助结节诊断和治疗至关重要。对于超声图像，卷积神经网络（CNNs）和 Transformer，广泛应用于自然图像，无法获得满意的分割结果，因为它们要么无法获得精确的边界，要么分割小物体。
为了解决这些问题，我们提出了一种新的边界保持组装 Transformer UNet（BPAT-UNet）用于超声甲状腺结节分割。在提出的网络中，设计了一个边界点监督模块（BPSM），采用两种新颖的自注意力池化方法，通过一种新颖的方法增强边界特征并生成理想的边界点。同时，构建了自适应多尺度特征融合模块（AMFFM），融合不同尺度的特征和通道信息。最后，为了充分整合高频局部和低频全局的特征，在网络的瓶颈处放置了组装 Transformer 模块（ATM）。通过将可变形特征和特征之间的相关性引入 AMFFM 和 ATM 的上述两个模块，对其进行特征之间的相关性进行了特征描述。作为设计目标并最终展示，BPSM 和 ATM 促进了所提出的 BPAT-UNet 进一步约束边界，而 AMFFM 则有助于检测小物体。
与其他经典分割网络相比，所提出的 BPAT-UNet 在可视化结果和评估指标上显示出优越的分割性能。在 TN3k 公共甲状腺数据集上，Dice 相似系数（DSC）为 81.64%，不对称 Hausdorff 距离（HD95）的第 95 个百分位数为 14.06，显著提高了分割精度；在我们的私有数据集上，DSC 为 85.63%，HD95 为 14.53。
本文提出了一种用于甲状腺超声图像分割的方法，该方法达到了较高的准确性，满足了临床需求。代码可在 https://github.com/ccjcv/BPAT-UNet 上获得。
