TransFSM: Fetal Anatomy Segmentation and Biometric Measurement in Ultrasound Images Using a Hybrid Transformer

被引:14
作者
Zhao, Lei [1 ]
Tan, Guanghua [1 ]
Pu, Bin [1 ]
Wu, Qianghui [1 ]
Ren, Hongliang [2 ]
Li, Kenli [1 ]
机构
[1] Hunan Univ, Coll Comp Sci & Elect Engn, Changsha 410082, Peoples R China
[2] Chinese Univ Hong Kong, Dept Elect Engn, Hong Kong 999077, Peoples R China
关键词
Ultrasonic variables measurement; Ultrasound image segmentation; fetal biometric measurements; hybrid transformer; deformable self-attention; U-NET; ARCHITECTURE; ATTENTION;
D O I
10.1109/JBHI.2023.3328954
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Biometric parameter measurements are powerful tools for evaluating a fetus's gestational age, growth pattern, and abnormalities in a 2D ultrasound. However, it is still challenging to measure fetal biometric parameters automatically due to the indiscriminate confusing factors, limited foreground-background contrast, variety of fetal anatomy shapes at different gestational ages, and blurry anatomical boundaries in ultrasound images. The performance of a standard CNN architecture is limited for these tasks due to the restricted receptive field. We propose a novel hybrid Transformer framework, TransFSM, to address fetal multi-anatomy segmentation and biometric measurement tasks. Unlike the vanilla Transformer based on a single-scale input, TransFSM has a deformable self-attention mechanism, so it can effectively process multi-scale information to segment fetal anatomy with irregular shapes and different sizes. We devised a boundary-aware decoder (BAD) to capture more intrinsic local details using boundary-wise prior knowledge, which compensates for the defects of the Transformer in extracting local features. In addition, a Transformer auxiliary segment head is designed to improve mask prediction by learning the semantic correspondence of the same pixel categories and feature discriminability among different pixel categories. Extensive experiments were conducted on clinical cases and benchmark datasets for anatomy segmentation and biometric measurement tasks. The experiment results indicate that our method achieves state-of-the-art performance in seven evaluation metrics compared with CNN-based, Transformer-based, and hybrid approaches. By knowledge distillation, the proposed TransFSM can create a more compact and efficient model with high deploying potential in resource-constrained scenarios. Our study serves as a unified framework for biometric estimation across multiple anatomical regions to monitor fetal growth in clinical practice.
引用
收藏
页码:285 / 296
页数:12
相关论文
共 40 条
[1]  
Babalola KO, 2008, LECT NOTES COMPUT SC, V5241, P409, DOI 10.1007/978-3-540-85988-8_49
[2]   STATISTICAL METHODS FOR ASSESSING AGREEMENT BETWEEN TWO METHODS OF CLINICAL MEASUREMENT [J].
BLAND, JM ;
ALTMAN, DG .
LANCET, 1986, 1 (8476) :307-310
[3]  
Cao H., 2021, arXiv
[4]   Diagnostic performance of third-trimester ultrasound for the prediction of late-onset fetal growth restriction: a systematic review and meta-analysis [J].
Caradeux, Javier ;
Martinez-Portilla, Raigam J. ;
Peguero, Anna ;
Sotiriadis, Alexandros ;
Figueras, Francesc .
AMERICAN JOURNAL OF OBSTETRICS AND GYNECOLOGY, 2019, 220 (05) :449-+
[5]   End-to-End Object Detection with Transformers [J].
Carion, Nicolas ;
Massa, Francisco ;
Synnaeve, Gabriel ;
Usunier, Nicolas ;
Kirillov, Alexander ;
Zagoruyko, Sergey .
COMPUTER VISION - ECCV 2020, PT I, 2020, 12346 :213-229
[6]   Performance measure characterization for evaluating neuroimage segmentation algorithms [J].
Chang, Herng-Hua ;
Zhuang, Audrey H. ;
Valentino, Daniel J. ;
Chu, Woei-Chyn .
NEUROIMAGE, 2009, 47 (01) :122-135
[7]   Rethinking the unpretentious U-net for medical ultrasound image segmentation [J].
Chen, Gongping ;
Li, Lei ;
Zhang, Jianxun ;
Dai, Yu .
PATTERN RECOGNITION, 2023, 142
[8]   AAU-Net: An Adaptive Attention U-Net for Breast Lesions Segmentation in Ultrasound Images [J].
Chen, Gongping ;
Li, Lei ;
Dai, Yu ;
Zhang, Jianxun ;
Yap, Moi Hoon .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2023, 42 (05) :1289-1300
[9]  
Chen J., 2021, arXiv
[10]   Automated ultrasound assessment of amniotic fluid index using deep learning [J].
Cho, Hyun Cheol ;
Sun, Siyu ;
Hyun, Chang Min ;
Kwon, Ja-Young ;
Kim, Bukweon ;
Park, Yejin ;
Seo, Jin Keun .
MEDICAL IMAGE ANALYSIS, 2021, 69